Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Study
No. 44 Role of Financial Frictions
in Monetary Policy Transmission in India
Different Types of Public Expenditure in India
Development Research Group
Department of Economic and Policy Research
Reserve Bank of India
Shesadri Banerjee
Harendra Behera
Sanjib Bordoloi
Rakesh Kumar
Issued for Discussion
DRG Studies Series
Development Research Group (DRG) has been constituted in Reserve Bank of India in its Department of Economic and Policy Research. Its objective is to undertake quick and effective policy-oriented research backed by strong analytical and empirical basis, on subjects of current interest. The DRG Studies are the outcome of collaborative efforts between experts from outside Reserve Bank of India and the pool of research talent within the Bank. These studies are released for wider circulation with a view to generating constructive discussion among the professional economists and policy makers.
Responsibility for the views expressed and for the accuracy of statements contained in the contributions rests with the author(s).
There is no objection to the material published herein being reproduced, provided an acknowledgement for the source is made.
DRG Studies are published in RBI web site only and no printed copies will be made available.
Director
Development Research Group
DRG Study
Role of Financial Frictions in
Monetary Policy Transmission in India
by
Shesadri Banerjee
Harendra Behera¤
Sanjib Bordoloi₸
Rakesh Kumar
The views expressed in this study are of the authors alone, and not of the institutions to which they belong. The usual disclaimer applies. * Madras Institute of Development Studies, 79 Second Main Road, Gandhinagar, Adyar, Chennai 600020, India. Tel: 91-44-24412589/24411574 (Ext: 340). E-mail: [email protected] ¤ Reserve Bank of India, Central Office Building, Shahid Bhagat Singh Marg, Mumbai 400001, India. E-mail: [email protected] ₸ Reserve Bank of India, Central Office Building, Shahid Bhagat Singh Marg, Mumbai 400001, India. E-mail: [email protected] Reserve Bank of India, Central Office Building, Shahid Bhagat Singh Marg, Mumbai 400001, India. E-mail: [email protected]
Acknowledgements
We are deeply grateful to Dr. Michael Debabrata Patra, Executive Director, Reserve
Bank of India (RBI), for his encouragement to pursue this Development Research
Group (DRG) study. We are thankful to Mr. Sitikantha Pattnaik for shaping the scope
of this research. We would like to thank Dr. Satyananda Sahoo, the former Director of
DRG in RBI, and his team members for their continuous support at different stages of
the study. The work has also been benefitted from valuable comments and
suggestions of Professor Chetan Ghate and Dr. Rudrani Bhattacharya who were the
discussants of the study presented at Annual Conference of Department of Economic
and Policy Research at Kochi (June 2017) and the comments of other participants at
the conference.
Executive Summary
In macroeconomics literature, the mechanism of monetary policy transmission (MPT)
has been a subject of extensive research in many countries over the last couple of
decades. In the course of such investigations, researchers have found financial market
frictions as one of the major determinants of propagation mechanism of the monetary
policy shocks. The credit channel based explanation of MPT attributes weak
transmission of monetary policy in emerging market and developing economies
(EMDEs) is due to the predominance of financial market frictions. Presence of
information asymmetries, limited enforceability of contracts and heterogeneity among
the economic agents give rise to frictions in the financial market transactions, which
play a crucial role in determining the degree of pass-through and speed of adjustments
in the MPT mechanism. In this study, we examine the critical role of different financial
frictions and the associated structural rigidities in the MPT in India.
At the outset, we document the stylised facts from the cyclical properties of the real
and financial variables, investigate the MPT mechanism using the Structural Vector
Autoregression (SVAR) methodology and explore the potential sources of financial
frictions that can deter the transmission process. First, the empirical regularities show
that: (i) the operational target (weighted average call money rate) and the policy
instrument (repo rate) are closely related; (ii) relatively strong co-movement exists
between the business cycle and the credit growth cycle; and (iii) counter-cyclical
movement of interest rate spread indicates linkage between the real and financial
sectors of the economy. Second, the SVAR analysis, based on seasonally adjusted
quarterly data of real output, consumer price inflation, non-food credit growth, deposit
interest rate, lending interest rate and weighted average call money rate (WACR) over
the sample period of 1999:Q4 to 2015:Q3, reveals slow and weak transmission of
monetary policy shocks through a combination of interest rate and bank lending
channels. The peak effects of a monetary policy shock, on an average, are observed
with a lag of three to four quarters. Moreover, the effects persist for nearly eight to
twelve quarters. Third, we find evidence from the literature on different types of
frictions prevailing in the bank dominated credit market in India. Broadly, these include
financially excluded segment of the population, credit-constrained households,
interest rate rigidity, policy-driven market distortions related to administered interest
rate on small savings, high statutory liquidity ratio (SLR) and capital adequacy ratio for
banks. In view of the empirical findings and evidence from the literature, we take this
study forward to understand the role of different frictions related to the credit market
structure and the banking sector of the economy.
i
We propose a New Keynesian Dynamic Stochastic General Equilibrium (NK-DSGE)
model with an imperfectly competitive banking sector at the core. Following Gerali et
al. (2010) and Anand et al. (2014), we develop a modelling framework using Indian
economy specific features of liquidity-constrained households, competitive labour
market and reserve requirements for the commercial banks. We incorporate a variety
of real, nominal and financial shocks to the prototype economy in order to pin down
the business cycle features and quantify the variance decomposition of shocks.
Combining the methods of calibration and Bayesian estimation for the sample period
1999:Q4 through 2015:Q3, the baseline parameterisation is configured and validated
with second order moments of the data.
Simulation of the baseline model replicates co-movement of the credit market interest
rates with incomplete pass-through and produces contractionary effects on the real
and financial variables as a consequence of a positive interest rate shock. Following
a positive interest rate shock, the spectrum of interest rates shifts up and squeezes
the demand for credit. This leads to contractionary effects emanating from the demand
side of the economy via reduction of consumption and investment demand, and from
the supply side via cost of physical capital. This two-pronged contraction leads to a
sharp decline in the demand for factors of production, in particular the labour, which
drives down the aggregate output and inflation subsequently. The variance
decomposition results show that the transmission of an interest rate shock to
aggregate demand and inflation is low and sensitive to the degree of financial market
frictions and structural composition of the credit market.
Focusing on the credit market friction parameters, we undertake counterfactual
experiments and evaluate the responsiveness of MPT using the accumulated effects
over a time horizon of eight quarters. In general, it is observed that MPT improves as
friction in the financial system diminishes. More specifically, based on the elasticity
measure, our results suggest that: (i) presence of liquidity-constrained and collateral-
constrained households poses major obstacles for the transmission; (ii) easing of the
collateral constraint and greater financial inclusion can enhance the degree of
transmission more than proportionately; (iii) interest rate rigidity on the lending side
and composition of savers and borrowers in the credit market have important
implications for the transmission mechanism; and (iv) rigidity in the deposit interest
rate does not appear to be a significant determinant of the weak MPT though its role
becomes prominent as the depositor base expands in the economy.
We undertake simulation experiments further on the central bank loss function with
respect to a set of alternative policy rules, which include the conventional form of
Taylor rule, asset price augmented Taylor rule and credit cycle augmented Taylor rule.
The results show that except the case of housing price augmented Taylor rule, the
ii
standard form of Taylor rule with forecast-based inflation and contemporaneous output
turns out as the optimal one for all the policy frameworks under study. In fact, adjusting
the policy interest rate to smooth out the credit cycle exacerbates volatility of inflation
and output. In addition, comparing three different policy frameworks for stabilising
inflation and output, we find that inflation stabilisation is the most desirable policy
option for the central bank as it minimises the welfare loss irrespective of the policy
rules. Overall, it appears that targeting financial stability through monetary policy rule
may not be appropriate for the purpose of economic stabilisation.
iii
Contents
1 Introduction 4
2 Background of Study 92.1 Nexus between Monetary Transmission and Financial Market Frictions . . . 9
2.1.1 Sources of Frictions in Financial Market Transactions . . . . . . . . . 10
2.1.2 Financial Friction: A Determinant of Transmission Mechanism . . . . 11
2.2 Indian Experience of Monetary Policy Transmission . . . . . . . . . . . . . . 12
2.2.1 Operating Procedure of Monetary Policy . . . . . . . . . . . . . . . . 12
2.2.2 Role of Bank Lending Channel . . . . . . . . . . . . . . . . . . . . . 14
2.2.3 Impediments in Pass-through of Monetary Policy . . . . . . . . . . . 15
2.3 Stylised Facts on Indian Macroeconomic and Financial Variables . . . . . . . 16
2.3.1 Cyclical Behaviour of Interest rates, Credit and Output . . . . . . . . 16
2.3.2 Evidence on Monetary Transmission from SVAR Analysis . . . . . . . 19
3 The Model 213.1 Description of the Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2 Household Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.1 Liquidity-constrained Household . . . . . . . . . . . . . . . . . . . . . 23
3.2.2 Patient Household . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.3 Impatient Household . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.4 Entrepreneur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.5 Competitive Labour Market . . . . . . . . . . . . . . . . . . . . . . . 28
3.3 Producers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.1 Monopolistically Competitive Retailer . . . . . . . . . . . . . . . . . 29
3.3.2 Capital Goods Producing Sector . . . . . . . . . . . . . . . . . . . . . 29
3.3.3 Housing Goods Producing Sector . . . . . . . . . . . . . . . . . . . . 30
3.4 Banking Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4.1 Retail Branch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4.2 Wholesale Branch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.5 Fiscal Authority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.6 Central Bank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.7 Resource Constraint and Aggregation . . . . . . . . . . . . . . . . . . . . . . 36
3.8 Forcing Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1
4 Quantitative Analysis 374.1 Baseline Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.1.1 Calibrated Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.1.2 Estimated Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.1.3 Model Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.1.4 Variance Decomposition Results . . . . . . . . . . . . . . . . . . . . . 45
4.2 Transmission Mechanism of Monetary Policy Shock . . . . . . . . . . . . . . 46
4.2.1 Transmission to Banking Sector . . . . . . . . . . . . . . . . . . . . . 48
4.2.2 Transmission to Real Sector . . . . . . . . . . . . . . . . . . . . . . . 49
4.2.3 Transmission to Inflation . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.3 Financial Frictions and Monetary Transmission: Evidence from Counterfac-
tual Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.4 Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5 Conclusion 57
6 Bibliography 60
7 Appendix A: Alternative Channels of Monetary Transmission 68
8 Appendix B: Econometric Specification of SVAR Analysis 70
9 Appendix C: Data Sources and Computations 71
10 Appendix D: Log-linearised Model and Description of Steady-state 7210.1 Liquidity-constrained Household . . . . . . . . . . . . . . . . . . . . . . . . . 72
10.2 Patient Household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
10.3 Impatient Household . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
10.4 Wholesale Goods Producing Entrepreneur . . . . . . . . . . . . . . . . . . . 73
10.5 Final Goods Producing Retailer . . . . . . . . . . . . . . . . . . . . . . . . . 74
10.6 Capital Goods and Housing Goods Producing Sectors . . . . . . . . . . . . . 74
10.7 Retail Banking Sector Operations . . . . . . . . . . . . . . . . . . . . . . . . 75
10.8 Wholesale Banking Sector Operations . . . . . . . . . . . . . . . . . . . . . . 75
10.9 Fiscal Policy Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
10.10Monetary Policy Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
10.11Aggregation & Market Clearing Conditions . . . . . . . . . . . . . . . . . . . 76
10.12List of Shock Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
10.13Description of Steady-state . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2
11 Appendix E: Figures on Counterfactual Experiments 78
3
1 Introduction
In the literature on macroeconomics, transmission mechanism of monetary policy remains
one of the fiercely debated areas. Monetary policy transmission (MPT) implies the process
through which policy action of the monetary authority is transmitted to the policy objec-
tives. Given the welfare consequences and distributional implications of the business cycle
fluctuations, the central bank needs to intervene from time to time with appropriate pol-
icy design. The success of such policy intervention depends on smooth functioning of the
transmission mechanism. Researchers have investigated the channels of MPT to explore the
effects of monetary policy shocks. Consequently, a host of studies have emerged following
different approaches explaining the pass-through mechanism of MPT to the real, nominal
and financial variables and describing the intensity and effi ciency of various channels.1
Although there is a consensus on the main conduits of MPT, determinants of the relative
strengths of different transmission channels are still not well established. Ambiguity arises
from the empirical evidence on the diverse nature of the transmission channels as well as
the temporal variations of MPT across countries over different sample periods.2 While
examining the role of a variety of structural and policy-driven factors, researchers have found
financial market frictions to be one of the major determinants of the impact and propagation
mechanism of monetary policy shocks in the economy. After the global financial crisis,
different economies have experienced sizeable decline in potential output and low inflation
expectations coexisting with weak household and corporate balance sheets. This kind of
experience emphasises the relevance and need for understanding the role of financial frictions
in the MPT mechanism.
Theoretically, it is argued that due to presence of information asymmetries, limited en-
forceability of contracts and heterogeneity among the economic agents, the financial market
is characterised by certain types of wedges, which can be viewed either in terms of price of
liquidity or the availability of liquid financial resources. In terms of price of liquidity, the
wedge can arise from the difference between the return received by providers of financial re-
sources and the cost of capital paid by capital users. In terms of the availability of liquidity,
the wedge can crop up from the difference between actual and desired liquidity of financial
resources accessed by the financially constrained market participants. Such wedges in the
financial market transactions are termed as financial frictions, which can influence the MPT
mechanism. Using a cross-country analysis, Cecchetti and Krause (2001) showed that the
transmission of a monetary policy action to the interest rate movements, domestic output
1A number of surveys on the theories of the monetary transmission mechanism are given in Bernanke(1993), Gertler and Gilchrist (1993), Kashyap and Stein (1993, 2000), Hubbard (1995), and Cecchetti (1995).
2See the cross-country analysis on MPT in Mishra and Montiel (2013), and Mishra et al. (2016).
4
and prices depend significantly on the structure of the country’s banking system and financial
markets. This lending view based on credit channel of MPT attributes the weaker trans-
mission mechanism of monetary policy in the emerging market and developing economies
(EMDEs) to their underdeveloped financial sector compared to the advanced countries.
Indian economy, similar to other EMDEs, features weak transmission mechanism of mon-
etary policy and incomplete pass-through due to its less deepened and fragmented finan-
cial market, costly intermediation and policy-driven market distortions. The transmission
process primarily works through the interest rate channel (RBI, 2005; Singh and Kalirajan,
2007; Patra and Kapur, 2010) and broad credit channel (Khundrakpam, 2011; Jain and
Khundrakpam, 2012). Aleem (2010) studied the credit channel, asset price channel and
exchange rate channel of MPT using VAR models for the period of 1996:Q4 to 2007:Q4 and
found the credit channel to be the only important channel of monetary transmission in India.
In our study, the cyclical properties of the real and financial variables reveal that: (i) the
operational target (weighted average call money rate) and the policy instrument (repo rate)
are closely related; (ii) strong co-movement exists between the business cycle and credit
growth cycle; and (iii) counter-cyclical movement of interest rate spread indicates some
degree of real and financial sector linkage in the economy. Using the seasonally adjusted
quarterly data on real output, consumer price inflation, non-food credit growth, deposit
interest rate, lending interest rate and weighted average call money rate for the sample
period of 1999:Q4 to 2015:Q3, we take a preview of the MPT in India based on a Structural
Vector Autoregression (SVAR) framework. Our empirical analysis suggests a slow and weak
transmission process through a combination of interest rate and bank lending channels.3 The
expected peak effect of a monetary policy shock, on an average, takes place with a lag of
three to four quarters. Overall, the effect persists for eight to twelve quarters.
The incomplete pass-through and long and uncertain time lag involved in MPT mech-
anism makes it diffi cult to predict the precise effects of monetary policy actions on the
economy. In order to improve MPT in the economy, the Reserve Bank of India (RBI) has
taken up different measures, which include changes in the operating framework, deregulation
of the interest rates, adoption of a more market-driven approach for evaluating the cost of
funds for the commercial banks like marginal cost based lending rate (MCLR) system and
others. Despite all the efforts of RBI on various occasions, the process of MPT does not work
seamlessly due to institutional bottlenecks and the structural arrangements of the bank-led
credit market (Patra and Kapur, 2010; Mohanty, 2016).
3This result is in line with Pandit and Vashisht (2011). They found that policy rate channel of transmissionmechanism, a hybrid of the traditional interest rate channel and credit channel, works in India as in otherEMDEs.
5
In India, the financial sector is largely dominated by the public sector commercial banks,
which formalise the credit market activities. These scheduled commercial banks play a
pivotal role in transmitting the policy-induced monetary impulses across different sectors of
the economy. Broadly speaking, the effects of MPT occurs in two steps: first, the change
in the policy rate affects the commercial bank interest rates, and second, the retail interest
rates of the banks impact the consumption/savings and investment decision-making of the
households and firms. Nevertheless, the expected outcome of policy intervention gets choked
off in these two steps due to several factors associated with the traits of bank-based formal
credit market and fiscal profligacy. Some of them are pointed out by Acharya (2017). First,
there exists a large segment of financially excluded population that can potentially deter
the transmission mechanism. Second, borrowers are often credit-constrained and can get
the credit subject to the value of their collaterals. Third, presence of the administered
interest rate structure on small savings constrains the adjustment of interest rate for the
deposits. Fourth, about 90 per cent of total liabilities of the commercial banks are in the
form of deposits, which are set at fixed interest rates. This discourages banks to reduce their
lending rates in line with the policy rate and imparts rigidity in the transmission process.
Fifth, persistence of the large market borrowing programme of the government hardens the
interest rate expectations. Besides, high statutory liquidity ratio, which provides a captive
market for government securities and suppresses the cost of borrowing for the government
artificially, partly strangles the MPT mechanism. Finally, the deterioration in banking sector
health due to low quality of assets and the unexpected loan losses in credit portfolios have led
to significant distortions in the pricing of assets. All these factors, in sum, lead to frictions
in the form of rigidities in the interest rate determination and cause impediments in the
pass-through of MPT to aggregate demand and inflation.
In view of the empirical observations and evidence from the existing literature on MPT
in India, we undertake this study to understand the role of frictions emanating from the
credit market composition and the banking sector. We propose a New Keynesian Dynamic
Stochastic General Equilibrium (NK-DSGE) model with an imperfectly competitive bank-
ing sector at the core. Following Gerali et al. (2010) and Anand et al. (2014), we de-
velop a modelling framework and augment the same by Indian economy-specific features like
liquidity-constrained household, competitive labour market and reserve requirements for the
commercial bank.
Our model consists of five building blocks: (i) household sector, (ii) production sector,
(iii) banking sector, (iv) fiscal authority, and (v) central bank. Household sector comprises
of heterogeneous agents. Primarily, there are two types of households: one is the liquidity-
constrained households excluded from the access to formal financial services, and other is
6
the financially included households. Again, the financially included group is characterised by
three different sub-groups, namely patient household (saver), impatient household (borrower)
and entrepreneur (borrower). These three types of households are different from each other
in terms of their time preferences (alternatively, the degree of impatience). Production sector
is operated by the perfectly competitive firms producing intermediate goods, capital goods,
housing goods and monopolistically competitive retail sector producing final goods. Banking
sector offers a one-period financial instrument like deposit contract (for the saver) and loan
contracts (for the borrowing household and firm). It operates with two branches, namely,
wholesale branch and retail branch. Wholesale branch operates competitively while retail
branch operates under a monopolistically competitive environment. Bank collects deposits
from the patient household and issues collateralised loans to the borrowing household and
the wholesale firm after meeting the statutory requirements in the forms of cash reserve ratio,
liquidity ratio and capital adequacy ratio. It accumulates capital from its profit. The fiscal
authority spends on final consumption goods and finances its spending by lump-sum taxes
and issuing government securities which are held by the commercial banks. The central bank
follows a Taylor-type interest rate rule by targeting the forecast-based inflation and current
state of business cycle.
The model features real frictions in the forms of external habit formation in consump-
tion, investment adjustment costs in the production of capital goods and housing goods.
Nominal friction is considered following Rotemberg (1982) in the price-setting behaviour of
retail goods sector. In the spirit of interest rate rigidity, credit-constrained borrowers and
regulatory norms as observed in India, financial frictions are modelled by the financially ex-
cluded population, collateral constraints, quadratic adjustment costs for interest rate setting
and maintaining the capital adequacy and reserve requirements of the bank. On one hand,
the collateral constraints for the borrowing household and firm, and on the other hand, the
balance sheet constraint and the law of motion for capital accumulation of the bank together
construct a built-in feedback mechanism between the real and financial sectors of the model.
Following the business cycle literature, we incorporate eight exogenous shocks in the model
namely total factor productivity, marginal effi ciency of investment, monetary policy, fiscal
spending, mark-up, preference for housing goods, loan to value (LTV) ratio for the borrowing
household and entrepreneur.
Baseline parameterisation of the model is configured by combining the methods of cali-
bration and estimation. The well known deep parameters and steady-state shares are cali-
brated while the economy-specific friction parameters and shock structure are estimated with
quarterly data (1999:Q1 to 2015:Q3) using Bayesian methodology. The baseline model is val-
idated with the second order moments of the data based on volatility and cross-correlations
7
of the key macroeconomic and financial variables. Simulation results of the baseline model
replicate the co-movement of the credit market interest rates with incomplete pass-through.
In response to a positive interest rate shock, the spectrum of interest rates shifts up and
squeezes the demand for credit. As a consequence, the contractionary effects set in from
the demand side via reduction of consumption and investment demand, and supply side via
cost of capital in the economy. This leads to a sharp decline in the demand for factors of
production, in particular for the labour market, which drives down the aggregate output and
inflation subsequently. The variance decomposition results show that the transmission of an
interest rate shock to aggregate demand and inflation is paltry and subject to the structural
attributes and degree of financial market frictions.
Focusing on the credit market friction parameters, we undertake the counterfactual ex-
periments and evaluate the responsiveness of MPT using the accumulated effects over a time
horizon of eight quarters. In general, it is observed that MPT improves as the friction in
the financial system diminishes. More specifically, based on the elasticity measure, our re-
sults suggest that: (i) presence of liquidity-constrained and collateral-constrained households
poses major obstacles for the transmission, (ii) easing of the collateral constraint and greater
financial inclusion can enhance the degree of transmission more than proportionately, (iii)
interest rate rigidity on the lending side and composition of saver and borrower in the credit
market have important implications for the transmission mechanism, and (iv) rigidity in the
deposit interest rate does not appear to be a significant one for the weak MPT, though its
role becomes prominent as the depositors’base (i.e., proportion of savers) expands in the
economy.
Further, our policy experiments using central bank loss function with respect to a set
of alternative policy rules show that except for the case of housing price augmented Taylor
rule, the standard form of the Taylor rule with forecast-based inflation and contemporaneous
output stands out as the optimal one for all policy frameworks under consideration. Hous-
ing price augmented Taylor rule performs marginally better than the standard form of the
Taylor rule. In contrast, adjusting policy interest rate to smooth out the credit cycle does
not seem to be useful. Moreover, based on central bank loss function, we compare three
different policy frameworks for stabilising output and inflation such as, (i) higher weightage
for inflation stabilisation relative to output stabilisation, (ii) higher weightage for output sta-
bilisation relative to inflation stabilisation, and (iii) equal weightage for inflation and output
stabilisation. It is found that higher weightage for inflation stabilisation relative to output
stabilisation is the most desirable policy option for the central bank as it minimises welfare
loss irrespective of the policy rules. Overall, it appears that targeting financial variables in
the monetary policy rule may not be appropriate for the purpose of economic stabilisation.
8
The rest of the paper is organised as follows. Section 2 provides the background of study.
Section 3 lays out the model. Section 4 reports the quantitative analysis with results from
the baseline model, discussion based on counterfactual experiments, and policy implications.
Section 5 concludes the study.
2 Background of Study
2.1 Nexus between Monetary Transmission and Financial Market
Frictions
MPT describes the sequence of actions through which policy-induced changes in the nominal
money stock or the short-term nominal interest rate impact aggregate demand and inflation
(Taylor, 1995; Woodford, 2003, and Ireland, 2008). The qualitative feature and quanti-
tative significance of transmission mechanism vary across countries and over time periods.
Different competing views are found in the literature on MPT according to the channels
of transmission.4 These include the money channel, interest rate channel, credit channel
comprising bank lending and balance sheet channels, exchange rate channel, asset price
channel and expectation channel.5 Although the literature provides some unanimous views
on these channels, their relative importance for an effective monetary transmission is still
contentious. The effectiveness of different transmission channels varies depending upon the
economic structure and financial conditions. Empirical evidence, for example, has shown
that interest rate channel is the most relevant one for advanced economies due to their well-
developed financial markets. In contrast, credit channel is the major conduit of transmission
in the EMDEs. In case of small open economies with flexible exchange rates, where the inter-
est rate channel is relatively weak, the exchange rate channel appears to be more crucial for
the transmission mechanism (Mohanty and Turner, 2008; Kletzer, 2012). Researchers have
examined various structural factors that can potentially determine the relative importance
of different channels of monetary transmission. On the whole, it is observed that the finan-
cial market frictions play a major role in determining the nature and degree of pass-through
of a monetary policy shock to the macroeconomic and financial variables. In the following
subsections, first, we unfold the sources of frictions in the financial architecture, and then
explore their role for different channels of MPT.
4We have provided a brief discussion on these channels of transmission in Appendix A.5Even though such transmission channels have their distinguishing effects on the real economy, there
are possibilities for interlinkages between the channels through which they may magnify or countervail theinfluence of each other in the transmission process.
9
2.1.1 Sources of Frictions in Financial Market Transactions
In an economy, if all agents are homogenous, financial resources remain liquid and flow to
the most profitable project or individual who values it most. However, in reality, it does
not happen due to market incompleteness and heterogeneity among the economic agents in
multiple dimensions. There is a limit to the feasible range of intertemporal and/or intratem-
poral trades of claims. It means that the agents are unable to postpone their spending or
insure themselves to smoothen their consumption and/or investment. In such a situation,
distribution of funds becomes important to determine the flow of funds and their allocation.
This typical feature of financial market has received considerable empirical support from
different economic regions based on the cyclical properties of financial markets over different
time periods.
The nature of market incompleteness and/or heterogeneity among the agents plays a crit-
ical role to determine the degree of financial frictions and their implications for the economy.
Incompleteness of the financial market can be exogenous or endogenous (Quadrini, 2011).
In case of exogenously induced market incompleteness, certain assets may not be traded in
the market. For endogenous market incompleteness, markets can remain incomplete as the
participants may not be willing to involve in certain trades due to problems of information
asymmetry and limited enforcement (Brunnermeier et al., 2012).
Information asymmetries limit the ability of the sellers (say, lenders) to force the buyers
(say, borrowers) to fulfil their financial obligations. In this case, the limit emerges from the
inability of the seller to observe the buyer’s action. For example, if the repayment depends
on the performance of the business and the performance depends on the unobservable work
effort, the borrower may have an incentive to choose the low level of work effort. Again, let
us consider a situation where seller of a financial contract (say, lender) can observe whether
the buyer (say, borrower) is obeying the contractual obligations. But, there is no instrument
available using which the seller can enforce the contractual obligations. This gives rise to
limited enforceability problem and leads to market incompleteness.
Along with information asymmetry and limited enforcement problems, heterogeneities
among the economic agents with respect to different dimensions, such as endowments, time
preference, risk aversion, productivity and belief, lead to presence of at least two groups
of agents. One group of agents is financially constrained (borrowers) and seeks external
funds, while the other group (lenders) provides at least some of the financial resources to
the first group. In consequence of such market incompleteness and interaction between
the heterogeneous agents, one can observe a wedge to exist in the financial market, either
in terms of price of liquidity (i.e., difference between the return received by providers of
financial capital and the cost of capital paid by capital users) or the availability of liquid
10
financial resources (i.e., difference between the actual and desired liquid financial resources
availed by the financially constrained market participants). This wedge, precisely, defines
the friction of financial market.
Since late 1970s, a body of literature started to evolve providing the theoretical justifi-
cations for financial frictions at the micro-level (Townsend, 1979; Stiglitz and Weiss, 1981;
Hart and Moore, 1994; Kiyotaki, 2011) and their macroeconomic implications (Bernanke
and Gertler, 1989; Carlstrom and Fuerst, 1997; Kiyotaki and Moore, 1997; Cooley et al.,
2004; Kiyotaki and Moore, 2008; Gertler and Kiyotaki, 2010; Mendoza, 2010; Jermann and
Quadrini, 2012). Researchers have also investigated empirical validations using the time-
series data (Bernanke et al., 1999; Heathcote et al., 2009; Curdia and Woodford, 2010;
Brzoza-Brzezina and Kolasa, 2013; Merola, 2015; Copaciu et al., 2015; Galvaoa et al., 2016;
Guerrieri and Iacoviello, 2017), cross-sectional data (Aysun et al., 2013; Mateju, 2013; Guer-
rieri and Iacoviello, 2017; Mian et al., 2017) and panel data (Bhaumik et al., 2011). In this
entire gamut of work, we are focusing on a particular segment of literature that has recog-
nised the pivotal role of financial market frictions in determining the strength of different
transmission channels of monetary policy.
2.1.2 Financial Friction: A Determinant of Transmission Mechanism
The role of financial friction was primarily identified for the (broad) credit channel of trans-
mission. The literature on external finance premium (Bernanke et al., 1999) as well as
collateralised debt (Kiyotaki and Moore, 1997) has recognised the role of friction in credit
channel of monetary transmission. Altering the external finance premium on borrowing (i.e.,
price of credit) or the valuation of collateral (i.e., credit limit), instruments of monetary pol-
icy can affect the wedge to move countercyclically, and create an additional impact on the
real variables beyond its standard effect through the cost of capital. In case of the balance
sheet channel of transmission, the effect of monetary policy on aggregate demand, which
works through the policy rate to retail interest rates, largely depends on the magnitude of
external finance premium, valuation of collateral and borrower’s access to credit (Aysun et
al., 2013; Gertler and Kiyotaki, 2010; Iacoviello, 2015; Guerrieri and Iacoviello, 2017). In
case of the bank lending channel, changes in bank’s loan supply induced by monetary policy
actions affect the real economy subject to the imperfect substitutability between deposits
and other sources of finance for the bank. Such a structure tends to amplify the propaga-
tion mechanism of the monetary policy shock (Gambacorta, 2008; Dib, 2010; Gambacorta
and Signoretti, 2014). For both of these channels, friction in the credit market creates an
enhancement mechanism for the effects of monetary policy shocks by changing the cost and
availability of credit in response to changes in interest rates and other policy instruments.
11
Some studies, which are aligned with the broad credit channel, argue for the role of
frictions with reference to bank capital and risk-taking channels of transmission mecha-
nism. Under the bank capital channel, the strength of bank’s balance sheet (instead of the
borrower’s balance sheet) is the main focus. Due to the presence of capital adequacy re-
quirement, which places a constraint on issuing new equities and ownership, bank’s balance
sheet position provides an additional leverage on the effects of monetary policy shock (Blum
and Hellwig, 1995; Van den Heuvel, 2002). Under the risk-taking channel, banks search for
a higher yield in response to an increase in its risk appetite (Rajan, 2005; Borio and Zhu,
2012). For both the channels, a higher degree of financial friction on bank’s capital entails
higher borrowing premium for the bank, which forces them to reduce their credit supply.
Using the bank-level data, several studies find evidence for the significance of financial con-
straints commonly proxied by bank size, liquidity and capitalisation (Kashyap and Stein,
1995 and 2000; Kishan and Opiela, 2000; Khwaja and Mian, 2008).
Apart from the credit channel, the role of financial frictions can be found for other trans-
mission channels too depending on the structural attributes and policy environment of the
underlying economy. Effectiveness of the interest rate channel critically depends on the
state of financial development, segmentation of the financial market, access for the market
participants and level of market distortions due to policy interventions. As an example, ad-
ministered interest rate regime by policy intervention, lack of financial deepening and a large
group of small borrowers weaken the interest rate channel in the emerging market economies.
On the front of policy environment, Altunbas et al. (2009) found that the dramatic increase
in securitisation activity in Europe had feebled the effi cacy of interest rate channel, as it
allowed greater access to liquidity without any expansion of the bank’s balance sheet, and
hence the ability to continue lending in the face of a tightening of the monetary policy. For
the exchange rate channel, the transaction cost for cross-border financial transactions and
the country’s risk premium act as the key sources of frictions and take a prominent role in the
context of small open emerging market economies. Finally, for the asset price channel, the
strength of transmission conspicuously revolves around the information-related frictions of
the asset markets (Iacoviello and Neri, 2010; Aysun et al., 2013; Gambacorta and Signoretti,
2014).
2.2 Indian Experience of Monetary Policy Transmission
2.2.1 Operating Procedure of Monetary Policy
In the pre-reform period (before 1990-91), monetary policy regime in India was subject to
widespread financial repressions, high fiscal deficit and administered interest rate regime.
12
It was operated under a framework of automatic monetisation where the RBI facilitated
financing of fiscal deficit as the debt manager of the government. The then banking sector
was characterised by a high share of government ownership without any competition from
private banks and heavily regulated from exposure to the financial markets. The banks used
to invest mostly in risk-free assets such as government securities. However, the scenario
started to change with implementation of reforms in the operating framework of monetary
policy during the late 1990s. These reform measures were aimed at the development of new
institutions and instruments in the financial market to bring in effi ciency in the financial
system and implement market-determined interest rate regime.
The phasing out of automatic monetisation through ad-hoc Treasury bills since 1997
provided space for MPT to work towards output and inflation stabilisation, as well as better
environment for fiscal-monetary coordination. This was an important step towards inde-
pendent monetary policy. With progressive dismantling of the administered interest rate
regime, the RBI reduced both Cash Reserve Ratio (CRR) and Statutory Liquidity Ratio
(SLR) substantially from the level prevailing in the early 1990s. This led to a significant
reduction in lending rates of the commercial banks. The creation of Securities and Exchange
Board of India (SEBI) was an important institutional development to regulate the financial
market.
In the money market segment, Liquidity Adjustment Facility (LAF) was introduced
for repo operation in order to provide short-term liquidity to the banks in exchange of
government securities. This reduced the short-term volatility in the call money rate and
helped in smoothening the interest rate channel of MPT mechanism. A pure inter-bank call
money market and framework for auction-based repo/reverse repo were set up subsequently
for short-term liquidity management as well as to improve the policy transmission. In order
to strengthen the interest rate channel of transmission further, the RBI commenced a new
operating procedure in May 2011 that recognised the weighted average overnight call money
rate as the operating target for monetary policy.
All the reform measures of operating procedures, in sum, encouraged the role of market
forces in determining the interest rates. In the backdrop of these developments, the RBI
sets its policy rate, which is repo rate, and thereby provides a signal to the economy at the
short-end level with an overall objective of influencing deposit and lending rates to impact
the output and prices.
The RBI uses its policy rate as an important counter-cyclical tool to achieve the policy
objectives of growth with price and financial stability. However, the effectiveness of mone-
tary policy critically depends on the strength of the transmission mechanism, which further
rests on the structural features of the real and financial sectors in the economy. The lit-
13
erature, which has evolved on the MPT mechanism in India, provides evidence for all the
channels of transmission mechanism mentioned in Section 2.1, but their relative strengths
vary significantly with different magnitudes and lags. Evidence can be found in Joshi, Sag-
gar and Ray (1998), Al-Mashat (2003), RBI (2005), Mohan (2008), Mohanty and Turner
(2008), Aleem (2010), Patra and Kapur (2010), Bhattacharya, Patnaik, and Shah (2010),
Mohanty (2012), Kapur and Behera (2012), and Singh (2012). These studies have examined
the monetary transmission mechanisms in India using different econometric methodologies
and sample periods.
2.2.2 Role of Bank Lending Channel
Out of the bulk of literature, a broad consensus has emerged that the bank lending channel
is the most impactful channel through which monetary policy can affect the macroeconomic
variables with a lag of two to three quarters. The bank lending channel works with an overlap
of the interest rate channel and credit channel. After a policy shock, deposit and lending
interest rates are adjusted by the commercial banks, which affect the borrower’s and lender’s
balance sheet, availability of credit, spending and investment decisions of households and
entrepreneurs and finally impact the output and inflation. Change in the policy interest rate
is an important determinant of household’s and firm’s demand for bank credit. Although it is
slow, but significant and robust. It entails incomplete pass-through of policy rate changes to
bank interest rates, and transmits to aggregate demand. Das (2015) has pointed out some of
the features of this particular channel in the following way. First, the extent of pass-through
to the deposit rate is larger and the adjustment is relatively faster than that to the lending
rate. Second, empirical evidence suggests an asymmetric adjustment to the monetary policy
shock. The lending rate adjusts more quickly to monetary tightening than to its loosening.
Third, the deposit rates do not adjust upwards in response to monetary tightening, but do
adjust downwards to its loosening.
In the Indian context, there are empirical studies showing the existence and significance
of the bank lending channel of the monetary policy. Using a VAR framework on quarterly
panel data of banks for the period 1997 to 2002, Pandit et al. (2006) found that the changes
in CRR and the bank rate get transmitted to the bank lending with the impact being much
stronger on small banks than large banks. Using the annual panel data of banks from
2000 to 2007, Bhaumik et al. (2010) examined the implications of bank ownership on the
transmission of monetary policy to the supply of bank credit. They observe that the bank
lending channel of MPT is more effective under the deficit than the surplus in liquidity
condition. Pandit and Vashisht (2011) examined the transmission of repo rate from the
perspective of demand for bank credit in India. Using monthly data from January 2001 to
14
August 2010 in a panel framework of seven emerging market economies including India, they
found that change in policy interest rate is an important determinant of firm’s demand for
bank credit. Considering a sample from the post LAF period (2001:Q3 to 2011:Q3) and
deploying an approach similar to Hendry’s general-to-specific method, Khundrakpam (2011)
found that policy rate induced expansion/contraction in deposit forces banks to adjust their
credit portfolio. The transmission of policy rate to real bank credit growth takes about
seven months over the full sample period as well as across various sub-sample periods. Over
the full sample period, a 100 basis points increase in policy rate was found to reduce the
annualised growth in real bank credit by 2.17 per cent. However, the magnitude of the
impact of policy interest rate on bank credit has been observed to decline during the post
Global Financial Crisis period. While pointing out such asymmetric adjustment behaviour
of deposit and lending interest rates, Sen Gupta and Sengupta (2014) and Das (2015) argued
for the predominance of the bank lending channel of monetary transmission in India.
2.2.3 Impediments in Pass-through of Monetary Policy
In India, MPT remains weak as a consequence of number of frictions originating from the
institutional framework, policy-driven market distortions and under-developed financial mar-
ket. A few of them are discussed below.
It is widely recognised that fiscal profligacy continued to remain intertwined with the
banking sector through reserve requirements of banks in the form of government securities,
which is known as statutory liquidity ratio or SLR. Lahiri and Patel (2016) showed that
policy-induced binding constraint on banks due to the SLR component can result in higher
lending rate spread in response to a reduction in policy rate. The rising spread leads to a
contraction in the economy instead of an expansionary traction.
Agricultural credit has a significant share of bank credit in India. Given the large share
of population depending on the agricultural sector, the government announced a "Compre-
hensive Credit Policy" in June 2004 with special focus on agricultural credit. Agricultural
credit increased from 13 per cent in 2001-02 to around 39 per cent in 2012-13. During the last
two-and-half decades, priority sector lending, Interest Subvention Scheme for crop loans and
loan waiver schemes in response to agrarian discourse have been implemented by the gov-
ernment. However, these are credit market distortions faced by banks having an important
bearing on the MPT mechanism. The credit market distortion in the form of priority sector
lending, which is deployed under policy rules, does not respond to either market signals or
subsequent monetary policy impulses. This restricts the impact of transmission of monetary
policy (Prasad & Ghosh, 2005).
Interest rates in India are largely determined by the market except in the case of small
15
savings instruments. The interest rates, which are offered on small savings on a medium-term
basis and changed quarterly, have been considered as one of the impediments for the MPT
mechanism. The small savings instruments such as National Savings Certificates, Provident
Funds and other postal savings are linked to interest rates of the government securities and
are generally higher than the deposit rates offered by banks. These instruments also pegged
with tax incentives to stimulate household savings. As a result, the administered nature of
small savings rates provides a competitive interest rates floor to the commercial bank. This
puts a binding constraint on the commercial banks to align their deposit rates close to small
savings rates and leads to higher adjustment cost for commercial banks. Accordingly, this
results in incomplete pass-through of changes in policy rate to deposit rate, with even lesser
degree on lending rates.
Further, the prevalence of oligopolistic market structure in the Indian banking sector
poses a bottleneck for the pass-through mechanism of MPT. Although the banking sector has
been opening up gradually over the last decade for the private and foreign banks, the degree
of market imperfection is quite prominent. The data of net interest margin (NIM), which is
used as an indicator of market competitiveness for the banking sector (Mandelmann, 2010),
shows that during the period 1992 to 2010, Indian banking sector became more competitive
but at a decreasing pace. The NIM declined by 18 percentage points between the period
1992-96 and 1997-2001. During 2002-07, it went down by 6 percentage points and after
2007, by 4.4 percentage points further. This indicates that the distortion due to market
imperfection is sizeable in the banking sector.6 Similar evidence can be found from the
interest rate spread adjustment behaviour of the commercial banks. Even after introducing
the marginal cost based lending rate system, commercial banks adjust the spread over the
marginal cost quite arbitrarily, which partly reflects the degree of credit market imperfection.
At times, these spreads are too large to be explained by the bank-level business strategy
and/or borrower-level credit risk premium (Acharya, 2017).
2.3 Stylised Facts on Indian Macroeconomic and Financial Vari-
ables
2.3.1 Cyclical Behaviour of Interest rates, Credit and Output
In view of the increasing role of interest rates as the policy instrument of monetary authority
in India, we have looked into the patterns in the movements of repo rate and call money rate
over the period of one-and-half decades using the quarterly data from 1999:Q4 to 2015:Q3.
6Claessens et al. (2001) find that an increased presence of foreign banks is associated with a reduction inprofitability and margins for domestic banks.
16
It is presented in Figure 1.
3
4
5
6
7
8
9
1 0
1 1
0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 1 0 1 1 1 2 1 3 1 4 1 5
CallM oneyRate Repo Rate
Figure 1: Patterns in Movements of Policy Instrument and Operational Target
Along with the co-movements of policy instrument (repo rate) and operational target (call
money rate), one can recognise two major policy-easing cycles that have been pursued by the
RBI during the period of study. The first one took place during the pre-crisis period (1999:Q4
to 2008:Q2) and the second one followed during the post-crisis period (2008:Q3 to 2015:Q3).
The sharp spikes during 2007-08 and upward movements of the interest rates from 2011-12
onwards are noticeable. These were the periods that witnessed supply-side driven double-
digit inflation, and necessitated policy tightening. Overall, a strong association between
the repo rate and call money rate is observed for the entire sample period with correlation
coeffi cient of 0.82, which is statistically significant at the level of 1 per cent. This association
becomes much stronger during the post-crisis period (0.95) compared to the pre-crisis period
(0.79).
17
. 0 8
. 0 6
. 0 4
. 0 2
. 0 0
. 0 2
. 0 4
. 0 6
. 0 8
. 0 4
. 0 3
. 0 2
. 0 1
. 0 0
. 0 1
. 0 2
. 0 3
. 0 4
0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 1 0 1 1 1 2 1 3 1 4 1 5
CreditCyclesBusiness_Cycles_G DPBusiness_Cycles_YG VA
Figure 2: Procyclical Movements in Credit Growth & Economic Activities
Table 1: Correlations between Cyclical Movements of Commercial Bank Credit and OutputBusiness Credit CycleCycle Pre-crisis Post-crisis Full SampleGDP 0.53** 0.62** 0.55**GVA 0.43** 0.47** 0.44**
While drawing up the patterns in the movements of policy instrument and its operational
target, it is imperative to examine the interaction between cyclical movements of the credit
expansion/contraction and the economic activities during the sample period. In Figure 2,
business cycle components of Gross Value Added (GVA), Gross Domestic Product (GDP)
and growth of aggregate credit provided by the scheduled commercial banks are plotted.
In Table 1, the cross-correlations between output and credit growth are reported. It is
found that growth of bank credit is procyclical and statistically significant, irrespective of
the measures of economic activities. Such procyclical behaviour of credit has intensified in
the post-crisis period than the pre-crisis period. Comparing Figures 1 and 2, further, it can
be noticed that the policy-easing cycles opted by the RBI were moderately followed by the
expansionary movements in the credit side and real activities in the economy.
In this context, we have explored the interaction between the synchronisation of credit
and real activities and the degree of financial friction over cyclical fluctuations. Following
Hall (2011), different measures of financial friction like interest rate spread between retail
lending and deposit rates, and term spread between the short-term and long-term government
bond yields are computed and their correlations with credit-to-GDP ratio and credit-to-GVA
18
ratio are examined. In Figure 3, we have presented the cyclical patterns of the spread and
credit-to-output ratio. In Table 2, the cross-correlation results are reported.
. 8
. 6
. 4
. 2
. 0
. 2
. 4
. 6
. 8
. 0 8
. 0 6
. 0 4
. 0 2
. 0 0
. 0 2
. 0 4
. 0 6
. 0 8
0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 1 0 1 1 1 2 1 3 1 4 1 5
Creditto G DP RatioInterestrate SpreadTerm spread
Figure 3: Countercyclical Movements of Spreads and Credit-to-GDP Ratio
Table 2: Correlations between Spreads and Credit-to-Output Ratio during Full SamplePeriod
Spread Credit-to-GDP Ratio Credit-to-GVA RatioLending/Deposit Interest Rate -0.40** -0.42**
Short/Long term Govt. Bond Yield -0.35** -0.34**
For all cases, it is observed that the measures of spread exhibit a counter-cyclical pat-
tern with the movements of credit-to-output ratio. This observation goes in line with the
prediction of the literature on macro-financial linkages. In the literature, it is argued that if
financial accelerator mechanism is in place, frictions will be low in the financial market during
the good times of business cycle and vice versa (Vlcek and Roger, 2012). The counter-cyclical
behaviour of different measures of spread confirms the same for Indian economy.
2.3.2 Evidence on Monetary Transmission from SVAR Analysis
In this sub-section, we take a preview of the MPT in India using Structural Vector Au-
toregression (SVAR) framework. SVAR has become a standard approach for evaluating the
effects of monetary policy shocks as it allows the modelling of recursive and non-recursive
structures of the economy with a parsimonious set of variables and facilitates the inter-
pretation of the contemporaneous correlation among the disturbances (Sousa and Zaghini,
19
2007). The econometric specification of SVAR representation and structural restrictions are
provided in Appendix B.
We use a six-variable SVAR specification, which includes real GDP growth (y), CPI
inflation (π), real non-food credit growth (b), interest rate on deposit(id), interest rate on
lending(ib)and call money rate (i) producing the monetary policy shock. The definitions
and sources of data are provided in Appendix C. The non-food credit is included to consider
the credit view of the policy transmission and is assumed to depend contemporaneously
on real income, inflation and the lending interest rate. Structural restrictions are imposed
for the identification of impact of monetary policy on output and prices. It is assumed
that the monetary policy variable does not respond to output and prices contemporaneously
because of lags in data release. It is also assumed that output and prices are not affected
contemporaneously by financial variables due to adjustment costs. Policy shocks do not
have an immediate impact on output and prices due to transmission lags. Deposit rates and
lending rates are assumed to depend on growth, inflation and credit demand of the economy.
Further, deposit rate is expected to impact the lending rate. The deposit rate responds
to economic activity, inflation and credit demand in the economy along with the lending
rate.7 The SVAR model is estimated using seasonally adjusted quarterly data for the period
1999:Q1 through 2015:Q3. The lag length of two is chosen based on the final prediction error
method to estimate SVAR. The impulse response of different variables to monetary policy
shock from the SVAR model is plotted in Figure 4.
.005
.004
.003
.002
.001
.000
.001
.002
1 2 3 4 5 6 7 8 9 10 11 12
Response of GDP growth
.4
.2
.0
.2
.4
1 2 3 4 5 6 7 8 9 10 11 12
Response of CPI inflation
.008
.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12
Response of Credit growth
.1
.0
.1
.2
.3
.4
.5
1 2 3 4 5 6 7 8 9 10 11 12
Response of deposit rate
.05
.00
.05
.10
.15
.20
.25
1 2 3 4 5 6 7 8 9 10 11 12
Response of lending rate
Impulse Responses to Monetary Policy Shocks
Figure 4: IRF Plots of a Monetary Policy Shock from SVAR
7According to our restrictions, the elements in the matrix representing impact of shocks on output andprices are assumed to be zero.
20
The impulse response plots of Figure 4 reveal the following observations. The deposit
and lending rates increase in response to increase in policy rate and the effect increases up
to the third quarter before tapering-off. Credit starts falling after two periods of lag and the
effects remain for a prolonged period. GDP growth declines in response to the policy rate
shock with a lag of two quarters and the effect dissipates slowly after peaking at the third
quarter. The negative impact on inflation follows after a decline in GDP growth and the
peak impact is observed with a lag of one quarter from the corresponding peak impact on
the GDP growth. However, a mild presence of price puzzle is observed as the inflation rate
increases up to second quarter in response to monetary policy shock.8 In sum, the result
provides evidence for both the interest rate and credit channels of monetary transmission to
macroeconomic and financial variables in the economy.
3 The Model
There has been a long-standing interest to incorporate financial friction in the mainstream
macroeconomic model, which has gained momentum in the post Global Financial Crisis
scenario. The associated literature offers two alternative approaches. The first strand of
research originated from the influential work of Bernanke and Gertler (1989), Carlstrom
and Fuerst (1997), and Bernanke et al. (1999), where financial friction is modelled by an
external finance premium that affects the price of credit in the economy via the financial
accelerator mechanism. The other seminal contribution came from Kiyotaki and Moore
(1997) and Iacoviello (2005), who introduced financial friction via collateralised debt, which
affects the quantity of credit availability. Both of these approaches are merged in the New
Keynesian Dynamic Stochastic General Equilibrium (NK-DSGE) models to examine the
effects of different structural and policy shocks emerging from the real and financial sectors
in a prototype economy.
The literature has been advanced by an explicit modelling of the role of financial inter-
mediaries into the analytical framework to provide a better understanding of the complex
interactions among the policy rate to the short-term market interest rates and the govern-
ment bond rates. Goodfriend and McCullam (2007) did pioneering work by introducing
banking sector in a DSGE model. They addressed interactions and differences between var-
ious types of interest rates based on the credit channel of their banking sector. Later, this
stream of research was enriched by the contributions from Curdia and Woodford (2016),
8Statistically significant response of the macroeconomic and financial variables are broadly observedaround the third quarter after a monetary policy shock except for inflation (as the confidence band of itsIRF includes both positive and negative quadrants).
21
Gertler and Kiyotaki (2009), Gertler and Karadi (2011), Andres and Arce (2012), Agenor et
al. (2013), Canzoneri et al. (2016) and Bhattarai et al. (2015).
In our study, we develop a medium-scale NK-DSGE model using the collateralised debt
approach. The model features imperfect credit market and financial intermediaries for an
emerging economy like India. This approach installs the broad credit channel of MPT and
highlights the role of borrowing and lending constraints as the key drivers of macro-financial
linkage in the economy. It is evident from the literature that credit channel via bank lending
is the principal route for MPT in India. Heavily regulated and imperfectly competitive
large bank dominated formal credit market with credit-constrained borrowers are the typical
features of the Indian financial sector. Such features motivate us to adopt a modelling
framework with the bank lending and balance sheet channels for the MPT mechanism.
3.1 Description of the Economy
We closely follow Gerali et al. (2010) and Anand et al. (2014) to build up our medium
scale DSGE model. The model is essentially an extension of the standard New Keynesian
framework with financially excluded population, savers, credit-constrained borrowers and
imperfectly competitive banking sector. A variety of frictions are modelled in the forms of
collateral constraints and symmetric adjustment costs except external habit formation in
consumption and inflation indexation in price-setting. Exogenous shocks are incorporated
as appropriate for the business cycle features of a developing economy. The environment of
our model is explained below.
The household sector consists of the representatives from the financially excluded, pa-
tient, impatient and entrepreneur groups. Financially excluded households are liquidity-
constrained and cannot participate in the financial market. In contrast, the representative
households from the patient, impatient, and entrepreneur groups are financially included but
heterogenous due to the difference in their time preference. Production side of the economy
comprises four sectors: (i) intermediate goods producing wholesale firms run by the entre-
preneurs, (ii) retailers who convert the intermediate goods into the final goods, (iii) capital
goods producing sector which produces new capital using old capital and investment and
(iv) housing goods producing sector that operates analogous to capital goods sector.
Operation of the representative commercial bank is managed by its two branches: whole-
sale branch and retail branch. The bank offers two types of one-period financial instruments:
one is deposit contract (for patient households) and the other is loan contract (for impatient
households and entrepreneurs). They collect financial resources via selling of deposit con-
tracts to the patient households; issue collateralised loans to the borrowing households and
22
the wholesale firms; meet the reserve requirements in the form of cash reserve ratio and
statutory liquidity ratio and macroprudential norm of the central bank in the form of capital
adequacy ratio; and accumulate capital from its profit. The balance sheet constraint of the
bank establishes the link between the business cycle and credit cycle in the economy through
bank capital. The degree of pass-through of the change in policy rate to retail deposit and
lending rates critically depends on the credit market imperfections, interest rate stickiness
and adjustment cost of bank’s capital-to-asset ratio.
There is a government that spends on final consumption goods. This fiscal expenditure
is financed by the lump-sum taxes and issuing of government securities that are held by the
commercial banks. The central bank follows a Taylor-type interest rate rule by targeting the
forecast-based inflation and current business cycle conditions.
3.2 Household Sector
The economy is populated by households and entrepreneurs, each one with a unit mass.
Households are segmented into two groups according to their access to the financial mar-
ket transactions. The first group is the liquidity-constrained households (R) that cannot
participate in the financial market. The other group of households actively participates in
the financial market operations and features heterogeneity with respect to their degree of
time preference. This financially included group consists of patient households (P ), impa-
tient households (I), and entrepreneurs (E). Patient households have a discount factor (βP )
which is higher than impatient households (βI) and entrepreneurs (βE) . Such a difference
in the time preference allows the patient households to be lenders and impatient households
and entrepreneurs to be borrowers in the model environment.
3.2.1 Liquidity-constrained Household
A representative ith household of the financially excluded segment of population consumes
the final goods CR,t(i) and supplies labour LR,t(i) to the packer in the competitive labour
market at real wage rate of wR,t. They maximise the following utility function:
UR,t =
[lnCR,t(i)−
L1+σlR,t (i)
1 + σl
](1)
subject to their budget constraint:
CR,t(i) ≤ wR,tLR,t (i) (2)
Hence, their optimal choice of consumption and labour supply yields:
23
1
CR,t(i)= λR,t (3)
LσlR,t (i) = wR,tλR,t (4)
where, λR,t is the Lagrangian multiplier implying the shadow price of consumption.
3.2.2 Patient Household
A representative patient household i chooses final consumption goods CP,t(i) subject to
habit formation on aggregate consumption, housing goods HP,t(i), labour supply LP,t (i),
and deposits Dt (i) in order to maximise the present value of life-time expected utility given
the periodical budget constraint. The expected utility function of a patient household is:
E0
∞∑t=0
βtP
[(1− σh) ln (CP,t(i)− σhCP,t−1) + εH,t lnHP,t(i)−
L1+σlP,t (i)
1 + σl
](5)
where, σh denotes the degree of habit persistence in consumption, σl is the inverse of
Frisch elasticity of labour supply, and εH,t is an exogenous shock to preference for housing
services. The flow of funds of the patient households is as follows:
CP,t(i) +Qht {HP,t(i)− (1− δh)HP,t−1(i)}+Dt (i) + TXP,t (i)
≤ wP,tLP,t (i) +
{(1 + idt−1
)πt
}Dt−1 (i) + Πr
P,t (6)
where, Qht is real price of housing, δh is depreciation rate of housing goods, wP,t is real
wage, idt is nominal interest rate on deposits, and πt is consumer price inflation at date t.
On the outflow of funds, expenditures are incurred for current consumption, accumulation of
housing goods, purchase of new deposit contracts, and lump-sum tax paid to the government
(TXP,t). On the inflow of fund, household receives labour income from the entrepreneurs,
interest income from the deposit holding of the previous period, and the profit received from
the ownership of retail goods producing firms(ΠrP,t
).
Patient household makes an optimal choice for {CP,t(i), HP,t(i), LP,t(i), Dt(i)}∞t=0 whichyields the following optimisation conditions:
(1− σh)(CP,t(i)− σhCP,t−1)
= λP,t (7)
24
[εH,t
HP,t(i)
]= λP,tQ
ht − βP (1− δh)λP,t+1Qh
t+1 (8)
LσlP,t (i) = wP,tλP,t (9)
λP,t = βP
(1 + idtπt+1
)λP,t+1 (10)
where, λP,t is Lagrangian multiplier for the budget constraint in real terms.
3.2.3 Impatient Household
The representative ith household from the impatient group derives utility from the consump-
tion of final goods CI,t(i) subject to habit formation on aggregate consumption, and housing
goods HI,t(i), and disutility from labour supply LI,t(i). It maximises the present value of
life-time expected utility:
E0
∞∑t=0
βtI
[(1− σh) ln (CI,t(i)− σhCI,t−1) + εH,t lnHI,t(i)−
L1+σlI,t (i)
1 + σl
](11)
subject to the sequence of budget constraint which is specified as:
CI,t(i) +Qht {HI,t(i)− (1− δh)HI,t−1(i)}+
{(1 + ibHt−1
)πt
}BH,t−1 (i) + TXI,t (i)
≤ wI,tLI,t (i) +BH,t (i) (12)
where, wI,t is real wage and ibHt is interest rate on borrowing at date t. Expenditures
are incurred for consumption, accumulation of housing goods, repayment of previous period
loans BH,t−1(i) with interest, and lump-sum tax payment to the government TXI,t(i). Inflow
of funds comes in the forms of labour income and current period borrowing.
In addition to the budget constraint, representative impatient household faces a borrow-
ing constraint that needs to be honoured to get loans from the bank. The household can
get credit upto the limit of expected nominal value of their collateral. Household uses its
accumulated physical assets of housing as the collateral. The borrowing constraint takes the
following form:
(1 + ibHt
)BH,t (i) ≤ εHLV,t (1− δh)Et
{Qht+1πt+1
}HI,t (i) (13)
25
where, εHLV,t is exogenously time varying LTV ratio for the borrowing households.
Impatient household optimally chooses {CI,t(i), HI,t(i), LI,t(i), BH,t(i)}∞t=0 which resultsinto the following optimal conditions:
(1− σh)(CI,t(i)− σhCI,t−1)
= λI,t (14)
[εH,tHI,t(i)
]= λI,tQ
ht − βI (1− δh)λI,t+1Qh
t+1 − εHLV,t (1− δh)µI,tQht+1 (15)
LσlI,t (i) = wI,tλI,t (16)
λI,t = βI
(1 + ibHtπt+1
)λI,t+1 +
(1 + ibHt
)µI,t (17)
where, λI,t and µI,t are the Lagrangian multipliers on the budget and borrowing constraints,
respectively.
3.2.4 Entrepreneur
There exists infinitely large number of entrepreneurs within a unit interval. The representa-
tive entrepreneur i derives utility from its final consumption (CE,t) subject to habit formation
on their aggregate consumption. The intertemporal discount factor of the entrepreneur is
denoted by βE. The present value of life-time expected utility function of the entrepreneur
is as follows:
E0
∞∑t=0
βtE [(1− σh) ln (CE,t(i)− σhCE,t−1)] (18)
The entrepreneur faces a budget constraint as well as a borrowing constraint which are
given below.
CE,t(i) + wR,tLR,t (i) + wP,tLP,t (i) + wI,tLI,t (i)
+
{(1 + ibEt−1
)πt
}BE,t−1 (i) +Qk
tKt (i) + ψt (ut)Kt−1 (i)
≤ YE,t (i)
Xt
+BE,t (i) +Qkt (1− δk)Kt−1 (i) (19)
(1 + ibEt
)BE,t (i) = εELV,t (1− δk)Et
{Qkt+1πt+1
}Kt (i) (20)
26
where,
YE,t = εA,t {ut (i)Kt−1 (i)}α L1−αt (i) (21)
Lt (i) = LγRR,t (i){LγP,t (i)L1−γI,t (i)
}1−γR (22)
ψt (ut) = ψa (ut − 1) +ψb2
(ut − 1)2 (23)
Xt =PtPE,t
(24)
In the above budget and borrowing constraints, ibEt is interest rate on borrowing from
bank for entrepreneurs, BE,t is amount of entrepreneurial borrowing, Qkt is real price of
physical capital, εA,t is the shock to total factor productivity, Lt is aggregate labour (after
combining the labour inputs from liquidity-constrained, patient and impatient households)
and Kt is physical capital used in wholesale goods production, ψt (ut) is cost of utilisation
of capital, YE,t is intermediate wholesale goods produced by the entrepreneur, and(1Xt
)is real marginal cost of wholesale goods production at date t. The share of capital in the
production function is α, the shares of labour of liquidity constrained, patient and impatient
households in the production are γR, {γ (1− γR)}, and {(1− γ) (1− γR)}, respectively andthe curvature parameters of the utilisation cost function are ψa and ψb.
In the entrepreneurial budget constraint of (19), expenditures are incurred for current
consumption, payment of wage bills to liquidity-constrained, patient, and impatient house-
holds for their labour supply, repayments of previous period’s debt, and utilisation cost of
capital. Entrepreneur receives inflow of resources in the form of output produced, borrowing
from the bank at current period, and selling of the undepreciated stock of physical capital
of the previous period.
The credit availability from the bank is determined by the stock of physical capital, which
is offered as collateral by the entrepreneurs. The loan restriction for entrepreneur is given by
equation (20). In the borrowing constraint, we have εELV,t which is exogenously time-varying
LTV ratio for the entrepreneur.
The sequences of {CE,t(i), Kt(i), ut, LR,t(i), LP,t(i), LI,t(i), BE,t(i)}∞t=0 are optimally cho-sen by the entrepreneur, and this results into the following optimal conditions:
(1− σh)(CE,t(i)− σhCE,t−1)
= λE,t (25)
27
λE,tQkt = εELV,t (1− δk)µE,tQk
t+1πt+1+
βEλE,t+1[rkt+1ut+1 + (1− δk)Qk
t+1 − ψt+1 (ut+1)]
(26)
rkt = ψa + ψb (ut − 1) (27)
wR,t = γRγ (1− α)
{YE,t (i)
Xt
}{1
LR,t (i)
}(28)
wP,t = (1− γR) γ (1− α)
{YE,t (i)
Xt
}{1
LP,t (i)
}(29)
wI,t = (1− γR) (1− γ) (1− α)
{YE,t (i)
Xt
}{1
LI,t (i)
}(30)
λE,t = βE
(1 + ibEtπt+1
)λE,t+1 +
(1 + ibEt
)µE,t (31)
where, marginal product of capital is: rkt = α{YE,t(i)
Xt
}{1
ut(i)Kt−1(i)
}; λE,t and µE,t are
Lagrangian multipliers on the budget and borrowing constraints of the entrepreneurs, re-
spectively.
3.2.5 Competitive Labour Market
Labour market is perfectly competitive where liquidity-constrained household, patient house-
hold and impatient household sell their labour to entrepreneur. All types of labour inputs
are bundled up via an aggregation technology by the entrepreneur in a costless way to pro-
duce homogenous labour input for the wholesale goods production. From the entrepreneur’s
choice of labour input, given the labour aggregator is in place, one can obtain the following
aggregate real wage weighted by share of different types of household’s labour in the labour
market.
wt = γwγRR,twγ(1−γR)P,t w
(1−γ)(1−γR)I,t (32)
where, γ =[γγRR {γ (1− γR)}γ(1−γR) {(1− γ) (1− γR)}(1−γ)(1−γR)
]−1Since there is no intra-group heterogeneity within the respective household group with
respect to their endowments, all individuals within a particular group face the same budget
constraint and objective function. Thus, they choose identical time paths for optimisation.
28
For this reason of symmetry within the group, hereafter we drop the household sector relevant
script i.
3.3 Producers
3.3.1 Monopolistically Competitive Retailer
The representative retailer buys homogenous intermediate goods at price PE,t from the en-
trepreneur, does the packaging with different brands at zero cost and turns them into differ-
entiated final goods. These differentiated final goods are sold at price Pt (j) in the imperfect
market that features monopolistic competition and nominal price rigidity. This price is in-
dexed by a weighted combination of last period inflation and steady-state level of inflation.
If the retailer adjusts the price of his goods beyond the indexation rule suggests, he will face
a quadratic adjustment cost parameterised by ϑp. Further, price of final goods is subject
to the mark-up shock due to the presence of exogenously time-varying price elasticity of
demand (εY,t) . The retail sector firm maximises:
E0
∞∑t=0
Λ0,t
[Pt (j)Yt (j)− PE,tYt (j)− ϑp
2
(Pt (j)
Pt−1 (j)− πθpt−1π1−θp
)2PtYt
](33)
subject to the sequence of demand constraints:
Yt (j) =
(Pt (j)
Pt
)−εY,tYt (34)
and finds the following optimal pricing condition for their goods:
1− εY,t +(εY,tXt
)−ϑp
[πt − πθpt−1π1−θp
]πt +βP
(λP,t+1λP,t
)ϑp(
πt+1 − πθpt π1−θp)πt+1
(Yt+1Yt
) = 0 (35)
3.3.2 Capital Goods Producing Sector
The capital goods producing sector is incorporated in order to derive the equation of the
market price of capital. This helps in determining the value of collateral of entrepreneurs
as they demand loans from the bank. In a perfectly competitive environment at the begin-
ning of each period t, these producers buy undepreciated last period’s capital stock of the
entrepreneurs (1− δk)Kt−1 at a price P kt . In addition, they purchase an amount of I
kt units
of the final goods from retailers at a price of Pt. The undepreciated capital of the previous
period is converted into the new capital at the rate of one-to-one. However, the final good
29
purchased from the retailers have this conversion subject to a quadratic adjustment costs.
Thus, the effective capital stock Kt, which is finally sold to entrepreneurs at a price P kt , has
its law of motion as given below:
Kt = (1− δk)Kt−1 +
[1− ϑk
2
{εik,t
(IktIkt−1
)− 1
}2]Ikt (36)
where, ϑk represents the adjustment cost of investment, εi,t is a shock to the productivity
of the investment and Qkt =
(PktPt
)is the price in real terms of the capital. As a result, the
capital producer maximises:
E0
∞∑t=0
Λ0,t[Qkt {Kt − (1− δk)Kt−1} − Ikt
](37)
subject to (36). Hence, the first order condition of optimisation of the capital goods produc-
ing firm turns out as:
Qt
1− ϑk2
{(εik,t
Ikt
Ikt−1
)− 1
}2− ϑk
{(εik,t
Ikt
Ikt−1
)− 1
}(εik,t
Ikt
Ikt−1
)+βPϑk
{(λP,t+1λP,t
)εik,t+1Qt+1
{(εik,t+1
Ikt+1
Ikt
)− 1
}(Ikt+1Ikt
)2} = 1 (38)
3.3.3 Housing Goods Producing Sector
Similar to capital goods producing sector, we add an explicit sector, which provides the basis
for market price of housing goods and subsequently, the valuation of collateral of impatient
household for taking loans from the bank. In this sector, firms operate in a competitive envi-
ronment and produces new housing goods using the previous period undepreciated housing
goods from borrowing households (1− δh)Ht−1 and Iht amount of final goods from the re-
tailers. Firms purchase undepreciated housing goods from borrowing households at price of
P ht and final goods from the retailers at Pt. While the old undepreciated housing goods can
be converted to new housing goods one-to-one, the new investment in house producing is
subject to quadratic adjustment cost. The law of motion of housing goods accumulation is
as follows:
Ht = (1− δh)Ht−1 +
[1− ϑh
2
{(IhtIht−1
)− 1
}2]Iht (39)
where, ϑh denotes the adjustment cost of investment in housing. The housing goods produc-
ing firms, therefore, maximise the following objective function:
30
E0
∞∑t=0
Λ0,t[Qht {Ht − (1− δh)Ht−1} − Iht
](40)
subject to (39). This optimisation exercise yields the following first order condition with real
price of housing goods Qht
(=
PhtPt
)as:
Qht
1− ϑh2
{(IhtIht−1
)− 1}2− ϑh
{(IhtIht−1
)− 1}(
IhtIht−1
)+βPϑh
{(λP,t+1λP,t
)Qht+1
{(Iht+1Iht
)− 1}(
Iht+1Iht
)2}= 1 (41)
3.4 Banking Sector
The representative bank j ∈ [0, 1] intermediates all financial transactions among the eco-
nomic agents in the model and works using two branches: one is the retail branch and the
other is the wholesale branch. The retail branch operates in a monopolistically competitive
environment through two departments. One department raises differentiated deposits from
the patient household, and the other department provides differentiated loans to the im-
patient household and wholesale goods producing entrepreneurs. The retail level branches
hold some market power in conducting their financial intermediation activity, which allows
them to set deposit interest rate and lending rates for the borrowing household and entre-
preneur. This type of banking structure enables us to examine different degrees of interest
rate pass-through from the change of policy rate, which can affect the real and nominal vari-
ables through the transmission mechanism. In contrast to the retail branch, the wholesale
unit - operating in a competitive market environment - provides wholesale loans and raises
wholesale deposits from the retail branches, and takes care of the position of bank capital.
3.4.1 Retail Branch
As in Gerali et al. (2010), we assume that units of deposit and loan contracts are differ-
entiated financial products bought by the households and entrepreneur, and are composed
by an aggregator with constant elasticities of substitution (CES). For a representative bank
j, the deposit contract to patient household, loan contract to impatient household and loan
contract to entrepreneur, elasticities of substitutions are εd, εbH , and εbE, respectively.
We assume that each patient household purchases a deposit contract from each single
bank in order to save one unit of her resource. On the other hand, each borrowing household
and entrepreneur purchases the loan contract from each single bank in order to meet their
ends. Such assumption goes with the standard Dixit-Stiglitz framework for imperfect market
structure which shows that the demand for an individual bank’s financial contract, either
31
deposit / loan, depends on the interest rate provided / charged by the bank relative to average
rates in the economy. Therefore, the demand functions for deposit and loan contracts for
the households and entrepreneur are given by:
Dt (j) =
(idt (j)
idt
)εdDt (42)
BH,t (j) =
(ibHt (j)
ibHt
)−εbHBH,t (43)
BE,t (j) =
(ibEt (j)
ibEt
)−εbEBE,t (44)
where, the average interest rates on deposit(idt), lending for household
(ibHt), and lending
for entrepreneur(ibEt)are defined as follows:
idt =
1∫0
idt (j)εd+1 dj
1
εd+1
(45)
ibHt =
1∫0
ibHt (j)1−εbH
dj
1
1−εbH
(46)
ibEt =
1∫0
ibEt (j)1−εbE
dj
1
1−εbE
(47)
Note that the aforementioned set of demand functions and the average interest rates for
the economy are derived from the expenditure minimisation exercise of the retail branches
of the representative bank.9
Retail Deposit Department: The retail deposit department of bank j collects patient
household’s deposits, Dt (j), and passes them to the wholesale unit, where deposits are
remunerated at rate of ist . The problem of the deposit unit is to maximise its expected
present value of profit after taking into account the quadratic adjustment cost parameterised
by φd. The optimisation problem can be written as:
9See the appendix of Gerali et al. (2010) for further details.
32
Maxidt E0
∞∑t=0
Λ0,t
istDt (j)− idtDt (j)−φd2
(idt (j)
idt−1(j)− 1)2Dt
s.t. Dt (j) =
(idt (j)
idt
)εdDt (48)
The first order condition of the above problem produces the following expression for
optimal deposit interest rate after imposing the symmetric equilibrium condition:
idt =
(εd
εd + 1
)ist −
(φd
εd + 1
)[(idtidt−1
)− 1
](idtidt−1
)+ (49)
βP
(φd
εd + 1
)[(idt+1idt
)− 1
](idt+1idt
)(Dt+1
Dt
)Retail Loan Department: Retail loan department of the bank optimally set the lending
rates for impatient household’s and entrepreneur’s borrowing in order to maximise its ex-
pected present value of profit and passes them to the wholesale branch at a uniform compet-
itive loan rate of ibt . Similar to the deposit department, loan department also faces quadratic
adjustment costs while changing the loan interest rates for household (parameterised by φbH)
and entrepreneur (parameterised by φbH). So, the retail loan unit maximises:
E0
∞∑t=0
Λ0,t
ibHt (j)BH,t (j) + ibEt (j)BE,t (j)− ibtBt (j)−φbH2
(ibHt (j)
ibHt−1(j)− 1)2BH,t − φbE
2
(ibEt (j)
ibEt−1(j)− 1)2BE,t
(50)
subject to (43)and (44), where
Bt (j) = BH,t (j) +BE,t (j) (51)
After imposing the condition of symmetric equilibrium, the optimal retail loan rates for
the household and entrepreneur become as follows:
ibHt =
(εbH
εbH − 1
)ibt −
(φbH
εbH − 1
)[(ibHtibHt−1
)− 1
](ibHtibHt−1
)+
βP
(φbH
εbH − 1
)[(ibHt+1ibHt
)− 1
](ibHt+1ibHt
)(BH,t+1
BH,t
)(52)
33
ibEt =
(εbE
εbE − 1
)ibt −
(φbE
εbE − 1
)[(ibEtibEt−1
)− 1
](ibEtibEt−1
)+
βP
(φbE
εbE − 1
)[(ibEt+1ibEt
)− 1
](ibEt+1ibEt
)(BE,t+1
BE,t
)(53)
3.4.2 Wholesale Branch
Wholesale branch collects deposits from the retail deposit department, generates loans from
the deposits and passes them to retail loan department. However, before converting the
financial resources from deposits into loans, the branch has to meet the reserve requirements
as stipulated by the RBI. Two types of reserve requirements are mandated, one is CRR
(parameterised by αc) and the other is SLR (parameterised by αs). CRR is the portion of
deposit that the bank is required to keep with the RBI in the form of cash. SLR is the
portion of bank’s deposit to be held in the form of liquid government securities. The RBI
varies these requirements to control credit supply by changing the availability of resources
available with the bank to make loans (Anand et al. 2014). The wholesale branch has access
to the interbank market to raise loan BIBt . Combining net worth of the bank Zt with the
interbank loan and deposit, the wholesale branch generates wholesale loan of Bt. Hence, the
balance sheet identity that the wholesale branch has to obey is as follows:
Bt (j) = (1− αc − αs)Dt (j) +BIBt (j) + Zt (j) (54)
We assume that capital stock of bank j is accumulated each period by adding up its
periodical earnings according to:
πtZt (j) = (1− δb)Zt−1 (j) + Πbt−1 (j) (55)
where, overall bank profit(Πbt−1)in the previous period made by the two branches of bank
j, and δb measures the resources used in managing bank capital and conducting overall
banking intermediation activity. Since we assume that bank capital is accumulated out of
its periodical earnings, the model has an in-built feedback mechanism between the real and
the financial side of the economy on the face of exogenous shocks.10
Further, we assume that there is a capital adequacy norm imposed by the central bank,
which sets a requirement for the representative commercial bank to maintain their capital
10If there is any adverse shock which deteriorates the macroeconomic conditions, banks profits will reducewhich will further weaken their ability to create new capital. Depending on the nature and size of theshock, it may result in the reduction of amount of loans supplied by the bank and exacerbate the originalcontraction.
34
to asset ratio[Zt(j)Bt(j)
]at level of κb. The bank is subject to a quadratic adjustment cost for
any deviation of its capital to asset ratio from the stipulated level. This modelling strategy
helps addressing the role of macroprudential norm in the bank capital channel of monetary
transmission.
The problem for wholesale branch is to choose loan Bt (j), deposit Dt (j), and interbank
borrowing BIBt (j) , so as to maximise the expected present value of profits subject to the
balance sheet constraint given by equation (54) and the law of motion for bank’s net worth
equation (55). Hence the wholesale branch will maximise:
E0
∞∑t=0
Λ0,t
ibtBt (j) + igt {αsDt (j)} − istDt (j)− itBIBt (j)
−Zt (j)− φz2
(Zt(j)Bt(j)
− κb)2Zt (j)
(56)
where, ibt , igt , i
st , and it are wholesale interest rate on loan, interest rate received from holding
of government bonds as SLR, wholesale deposit rate and interest rate for interbank loan,
respectively. The adjustment cost of bank capital is parameterised by φz. In a symmetric
equilibrium, the first order condition gives the following results:
ist = (1− αc − αs) it + αsigt (57)
ibt = it − φz(ZtBt
− κb)(
ZtBt
)2(58)
The above optimal conditions link the wholesale deposit and lending rates to the policy
rate it, interest rate on government bond, reserve requirements, and the leverage of the
banking sector. Finally, the profit of bank j, comes as the sum of earnings from the wholesale
and the retail branches and can be written by:
Πbt (j) = ibHt (j)BH,t (j) + ibEt (j)BE,t (j)− idtDt (j)− itBIB
t (j)− φd2
(idt (j)
idt−1 (j)− 1
)2Dt−
φbH2
(ibHt (j)
ibHt−1 (j)− 1
)2BH,t −
φbE2
(ibEt (j)
ibEt−1 (j)− 1
)2BE,t −
φz2
(Zt (j)
Bt (j)− κb
)2Zt (j)
(59)
3.5 Fiscal Authority
The government consumes an exogenously specified stream of spending Gt of final consump-
tion goods and finances this by lump-sum taxes of (TXP,t + TXI,t) and issuing bonds to the
bank through SLR. The government budget constraint is given by:
35
Gt +
(1 + igt−1πt
){αsDt−1 (j)} = (TXP,t + TXI,t) + {αsDt (j)} (60)
3.6 Central Bank
The central bank sets an interest rate rule (it) that follows a standard Taylor rule in the
short-run and is specified as given below:
(iti
)=
(it−1i
)φi {(πt+1π
)φπ (YtY
)φy}(1−φi)exp {εm,t} (61)
where, φi is the interest rate smoothing parameter, φπ and φy are the policy responses to
deviation of expected inflation πt+1 and output from their respective steady-state level. εm,tis the monetary policy shock.
3.7 Resource Constraint and Aggregation
The following resource constraint represents the final goods market equilibrium condition:
Yt = Ct + Ikt + Iht + Gt + ψt (ut)Kt−1 + δb
(Zt−1πt
)+ Adjt (62)
where, the aggregate consumption is:
Ct = CR,t + CP,t + CI,t + CE,t (63)
and, Adjt includes all types of adjustment costs incorporated in the model.
Physical asset in the form of housing good is aggregated as:
Ht = HP,t +HI,t (64)
Finally, credit (Bt) provided by the bank to the borrowing household and firm is aggre-
gated as:
Bt = BH,t +BE,t (65)
3.8 Forcing Processes
We have eight exogenous variables in our model and they follow AR(1) process as given
below:
36
(εA,tεA
)=
(εA,t−1εA
)ρAexp
{ξA,t}
(66)
(εik,tεik
)=
(εik,t−1εik
)ρik
exp{ξik,t}
(67)
(Gt
G
)=
(Gt−1
G
)ρgexp
{ξg,t}
(68)
(εm,tεm
)=
(εm,t−1εm
)ρmexp
{ξm,t}
(69)
(εy,tεy
)=
(εy,t−1εy
)ρyexp
{ξy,t}
(70)
(εH,tεH
)=
(εH,t−1εH
)ρhexp
{ξh,t}
(71)
(εHLV,tεHLV
)=
(εHLV,t−1εHLV
)ρHLV
exp{ξHLV,t
}(72)
(εELV,tεELV
)=
(εELV,t−1εELV
)ρELV
exp{ξELV,t
}(73)
The above shock variables drive the aggregate dynamics of our model.
4 Quantitative Analysis
We log-linearise the non-linear structure of decision rules, market clearing conditions and
resource constraints around the steady-state and obtain a short-run equation system. Our
quantitative analysis is premised on this log-linearised system of equations.11 In this section,
first, we set up the baseline parameterisation that works as a benchmark for rest of the
analysis, illustrations and discussions. Second, we explain the transmission mechanism of
monetary policy shock based on the properties of impulse response functions (IRF). Third,
we conduct the sensitivity experiments with respect to different financial friction parameters
and examine their role in the MPT mechanism. Finally, we discuss the policy implications
of our results for the betterment of transmission process and economic stabilisation.
11The system of equations and the steady-state specifications are laid out in Appendix D.
37
4.1 Baseline Model
We construct the baseline parameterisation of the model by synthesising the methods of cali-
bration and estimation. We calibrate some of the model parameters which are available from
the existing studies and macroeconomic time series data. In contrast, the parameters that
are more country-specific in nature like the heterogeneity in household sector composition,
frictions in the banking sector and persistence coeffi cients and standard errors of the exoge-
nous shocks, are estimated using the quarterly data of Indian macroeconomic and financial
variables over the sample period of 1999:Q4 to 2015:Q3.12 We deploy the methodology of
Bayesian rule as it allows the prior information to identify the parameters and impact of
shocks using the cross-equation restrictions given the general equilibrium set-up. We blend
the posterior means of the estimated parameters along with the well known calibrated pa-
rameters to create a baseline model for the Indian economy. Using this baseline model, we
study the impulse response properties of the monetary policy shock, which is the central
focus of this study.
4.1.1 Calibrated Parameters
We fall back on the existing DSGE literature to calibrate some of the structural parameters
and time-series data of relevant macroeconomic variables to pin down the steady-state shares.
In Table 3, the numerical values for calibration are provided. The proportion of liquidity-
constrained households (γR) is taken as 40 per cent according to the estimate of Gabriel et
al. (2011) for India. The heterogenous discount rates for patient household (βP ), impatient
household (βI) and entrepreneur (βE) are fixed at 0.96, 0.95 and 0.92, respectively based on
the average interest rates on deposits (8 per cent), households’borrowing (9.5 per cent) and
firms’borrowing (13 per cent) during the sample period of our study.
Table 3: Calibrated Parameters of Baseline Modelα βP βI βE δh δk σl εy εHLV εELV γR0.25 0.96 0.95 0.92 1.25% 2.5% 0.25 7 0.55 0.25 0.4
π ϑp θp εd εbH εbE αc αs κb C/Y G/Y Ik/Y D/Y BE/B4% 118 0.55 13 2.5 5.4 5.5% 21.5% 0.1 0.54 0.1 0.21 0.52 0.65
The degree of external habit formation (σh) is set to 0.66 similar to Banerjee and Basu
(2017). The Frisch elasticity of labour supply (σl) is taken as 0.25 in line with the elastic
12Sources of all the time series data are discussed in Appendix C.
38
nature of labour supply in India. The share of capital in production is set at 0.25 following
Gerali et al. (2010). The quarterly depreciation rates of physical capital (δk) and housing
goods (δh) are set to 2.5 and 1.25 per cent, respectively. The curvature parameters (ψa, ψb)
of the utilisation cost of capital are set following Silva et al. (2012).
Nominal friction in the form of price adjustment cost (ϑp) is chosen to be 118 following
Anand et al. (2010). The degree of inflation indexation is set to 0.55 (Sahu, 2013). Following
the Indian experience, capital adequacy requirement (κb), statutory liquidity ratio (αs) and
cash reserve ratio (αc) are set at 10, 21.5 and 5.5 per cent, respectively. The depreciation
rate of bank capital (δb) is chosen as 3.7 per cent (Anand et al., 2014). The elasticities of
demand for borrowing by the impatient household(εbH)and firm
(εbE), and for deposit
contract by the patient household(εd)are chosen in line with Silva et al. (2012).
The steady-state shares for consumption to GDP(CY
), gross capital formation to GDP(
Ik
Y
), government spending to GDP
(GY
), deposit to GDP
(DY
), entrepreneurial borrowing
to total borrowing(BEB
)ratio - are chosen based on their average value from quarterly data
for the sample period of study. Given the level of inflation target, the steady-state value of
inflation is chosen as 4 per cent. The long-run value of policy rate is taken as 7 per cent
based on the data of repo rate.
Steady-state values of technology and policy shocks are normalised to one. The steady-
state value of preference shock to housing is chosen to be 0.2 as proposed by Silva et al.
(2012). Following Gerali et al. (2010), we choose the steady-state values of LTV ratio as 0.55
and 0.25 for the impatient household(εHLV)and wholesale entrepreneur
(εELV), respectively.
4.1.2 Estimated Parameters
There are two constituents for implementing the Bayesian estimation for the unknown set
of parameters: one is the historical data for a set of observables and the other is prior
distributions. We consider the historical data series of real output (y), real gross capital
formation (ik), real fiscal consumption (g), CPI inflation (pi), retail deposit rate (id), retail
loan rate (il) and call money rate (i). Except the series for interest rates and inflation, all
other series are made stationary by taking the first differences. The series of all observables
are plotted in Figure 5.
39
20 40 600
0.1
0.2y
20 40 600.5
0
0.5ik
20 40 601
0
1g
20 40 600
0.1
0.2pi
20 40 600.05
0.1id
20 40 600.1
0.12
0.14il
20 40 600
0.1
0.2i
Figure 5: Plots of Historical Data Series
Next, the prior distributions for the relevant parameters are specified. Following the lit-
erature, we propose the priors that would fit with the Indian data. In course of specifying the
priors for the estimable parameters, we declare their respective probability density functions.
Selection of the probability density functions for the priors are based on the theoretical im-
plications of the relevant parameters in the model and the evidence from extant studies. As
example, the beta distribution is used for the fraction parameters, while the inverse gamma
distribution is specified for the parameters with non-negativity constraints. Due to lack of
the estimated DSGE models with financial market frictions for the EMDEs, and for India
in particular, we have less information regarding the standard deviations of the prior dis-
tributions. Thus, we select higher standard deviations and allow the data to determine the
location of the relevant parameters. The choice of such higher standard deviation for the
prior’s distribution is in line with Gabriel et al. (2012).
We obtain the joint posterior distribution of the estimated parameters by following the
Markov Chain Monte Carlo-Metropolis-Hastings (MCMC-MH) algorithm. This algorithm
simulates the smoothed histogram that approximates to the posterior distributions from
the prior distributions for the parameters of our interest. Two parallel chains are used
in the MCMC-MH algorithm. The univariate and multivariate diagnostic statistics show
convergence by comparing the ‘between’and ‘within’moments of multiple chains (Brooks
and Gelman, 1998).
In Table 4, the symbols and corresponding description of the estimated parameters (of
Figures 6 to 8) are listed. In Table 5, the prior and posterior means of the estimated
40
parameters are presented. The posterior means of estimated parameters are reported with
90 per cent confidence intervals subject to the posterior standard deviation. Figures 6 to
8 plot the prior versus posterior distributions. Figure 9 plots the diagnostic statistics of
multivariate convergence.13 Our estimation results suggest that all the parameters are well
identified and the posteriors are generated based on the information extracted from the
observables.14 The modes of the posterior distributions are significantly different from the
prior distributions which suggest that the information is extracted reasonably well from the
data to compute the posterior means. Combining the calibrated and estimated parameters
of Tables 3 and 5, we constitute the baseline parameterisation of our model for the sample
period under study.
Table 4: Description of Estimated ParametersParam eters Description Param eters Description
γ gamma ρg rho_g
ϑh rho_var_h ρy rho_epsy
ϑk rho_var_k ρh rho_epsh
φd phi_ id ρHLV rho_epslv_h
φbH phi_ ibh ρELV rho_epslv_e
φbE phi_ ib e σa SE_zeta_epsa
φz phi_ z σik SE_zeta_epsik
ϕi phi_ i σm SE_zeta_epsi
ϕπ phi_pi σg SE_zeta_g
ϕy phi_y σy SE_zeta_epsy
ρa rho_epsa σh SE_zeta_epsh
ρik rho_epsik σHLV SE_zeta_epslv_h
ρm rho_epsi σELV SE_zeta_epslv_e
13All the quantitative simulation and estimation of the model are done by using Dynare version 4.5.1. Fordetails of the computation procedure, see Dynare User Guide, 2013, Ch. 8.14Following the criteria of asymptotic information matrix and collinearity patterns of the parameters,
as suggested by Iskrev (2010a, b) and Iskrev and Rotto (2010a, b), identification of the shocks and thestructural parameters are examined. The strength of identification is verified by visual inspection of theplots of asymptotic information matrix.
41
0.1 0.3 0.50
100
200SE_zeta_epsa
1 2 30
10
SE_zeta_epsik
0.1 0.3 0.50
200
SE_zeta_epsi
0.1 0.3 0.50
10
SE_zeta_g
0 0.5 10
10
SE_zeta_epsy
0 0.2 0.4 0.60
10
SE_zeta_epsh
0.1 0.3 0.50
10
SE_zeta_epslv_h
0.1 0.3 0.50
50SE_zeta_epslv_e
0 0.4 0.80
5
gamma
Figure 6: Priors and Posteriors
4 60
0.5
1
rho_var_h
2 40
0.5
1
rho_var_k
0 20 400
0.1
phi_id
0 20 400
0.1
phi_ibh
0 20 400
0.1
phi_ibe
2 4 6 80
0.5
phi_z
0.8 10
10
phi_i
1.3 1.5 1.70
5
10phi_pi
0 0.2 0.40
5
10phi_y
Figure 7: Priors and Posteriors
0.5 10
10
20rho_epsa
0.2 0.4 0.6 0.80
5
rho_epsik
0.5 10
10
rho_epslv_h
0.2 0.6 10
5
rho_epslv_e
0 0.5 10
2
4
rho_epsy
0 0.5 10
1
2
rho_epsh
0 0.4 0.80
5
rho_g
0 0.5 10
2
rho_epsi
Figure 8: Priors and Posteriors
42
10 41 2 3 4 5 6 7 8 9 10
5
10
15Interval
10 41 2 3 4 5 6 7 8 9 10
0
10
20m2
10 41 2 3 4 5 6 7 8 9 10
0
100
200m3
Figure 9: Plots of Multivariate Convergence
Table 5: Prior Densities and Posterior Estimates of Baseline ModelEstim ated Priors Posteriors Estim ated Priors Posteriors
Param eters M ean D istribution M ean 90% C .I. Param eters M ean D istribution M ean 90% C .I.
γ 0.50 Beta 0.33 [0 .21, 0 .45] ρg 0.50 Beta 0.35 [0 .26, 0 .43]
ϑh 4 Gamma 4.09 [3 .57, 4 .57] ρy 0.50 Beta 0.49 [0 .34, 0 .65]
ϑk 4 Gamma 3.70 [3 .23, 4 .19] ρh 0.50 Beta 0.52 [0 .27, 0 .78]
φd 10 Gamma 13.45 [6 .88, 19.53] ρHLV 0.50 Beta 0.92 [0 .88, 0 .96]
φbH 10 Gamma 13.06 [6 .98, 18.49] ρELV 0.50 Beta 0.85 [0 .76, 0 .94]
φbE 10 Gamma 21.12 [13.04, 28.58] σa 0.10 Inv. Gamma 0.022 [0 .019, 0 .026]
φz 5 Gamma 4.85 [4 .07, 5 .62] σik 0.10 Inv. Gamma 1.616 [1 .238, 1 .960]
ϕi 0.80 Beta 0.86 [0 .82, 0 .91] σm 0.10 Inv. Gamma 0.014 [0 .012, 0 .016]
ϕπ 1.50 Normal 1.47 [1 .41, 1 .52] σg 0.10 Inv. Gamma 0.319 [0 .269, 0 .366]
ϕy 0.25 Normal 0.24 [0 .18, 0 .31] σy 0.10 Inv. Gamma 0.512 [0 .359, 0 .666]
ρa 0.50 Beta 0.91 [0 .87, 0 .94] σh 0.10 Inv. Gamma 0.113 [0 .023, 0 .278]
ρik 0.50 Beta 0.61 [0 .52, 0 .69] σHLV 0.10 Inv. Gamma 0.316 [0 .261, 0 .371]
ρm 0.50 Beta 0.33 [0 .15, 0 .51] σELV 0.10 Inv. Gamma 0.081 [0 .068, 0 .094]
4.1.3 Model Validation
In order to examine the reliability of baseline model, we compare the model generated
second order moments with their data counterpart for the major macroeconomic and financial
variables.15 In Table 6, descriptions of these key variables are provided.
15Given the property of stationarity of the model, the data series of output growth, consumption growth,investment growth and credit-to-output ratio are stationarised using Christiano-Fitzgerald asymmetric busi-ness cycle filter. The market interest rates, policy interest rate and CPI inflation rate are kept unchanged.Due to unavailability of long time series data related to the housing sector, we find a balanced sample startingfrom 2009: Q4 to 2015: Q3. Our model validation is, therefore, restricted within that time period.
43
Table 6: Description of Key VariablesVariab le Description Variab le Description Variab le Description
y Output qh Real price of housing u Capita l utilization
c Consumption mc Real marginal cost lp Labour Suply of Patient HH
ik Capita l investm ent pi CPI inflation lip Labour Suply of Impatient HH
ih Housing investm ent i Policy interest rate bh Household’s b orrow ing
b Borrow ing is Wholesa le dep osit rate be Entrepreneur’s b orrow ing
h Housing ib Wholesa le lend ing rate bib Interbank borrow ing
k Capita l id Retail dep osit rate rw Aggregate real wage
l Labour il Aggregate Retail lend ing rate ibh Retail lend ing rate for bhqk Real price of cap ita l pi_b Bank’s profit ibe Retail lend ing rate for be
The data and model comparison is done in two steps. First, we check the model generated
volatility with empirically observed volatility for a set of real and financial variables. Next, we
do the same for key cross-correlations among the macroeconomic and financial variables.16 In
Table 7, relative volatilities of four core variables —consumption to output, capital investment
to output, housing investment to output and bank credit to output —along with the volatility
of interest rate spread are presented based on the data and model. Although the model
slightly overpredicts, it comes close to the volatility indicators moderately.
Table 7: Comparing Volatilities between Data and ModelTarget Data Model(
σcσy
)1.27 1.15(
σik
σy
)3.03 3.26(
σih
σy
)1.55 2.46(
σbσy
)1.64 1.99
σ(il/id) 0.21 0.24
Further, we examine the model predictions for interrelationship among the major macro-
economic and financial variables in Table 8. It is noticeable that the model is predicting signs
16Note that, for the convenience of illustration and analysis, we have defined a new variable il as theweighted sum of retail lending rates to household
(ibH)and firm
(ibE). The weights are assigned based on
the steady-state shares of credit to household(BH
B
)and credit to firm
(BE
B
). il represents the economy-wide
average retail lending rate of the bank.
44
Table 8: Comparing Cross-correlations between Data and ModelCorrelations Data Model Correlations Data Model
(y, c) 0.30 0.43(ik, π
)0.28 0.19
(y, b) 0.20 0.12(ik, i)
-0.56 -0.14(y, ih
)0.46 0.50
(ih, i)
-0.37 -0.59(y, π) -0.29 -0.34
(ih, id
)-0.31 -0.70
(y, i) -0.64 -0.25(ih, il
)-0.52 -0.61(
b/y, il/id)
-0.52 -0.54(i, id)
0.90 0.60(b/y, π) -0.27 -0.38
(i, il)
0.80 0.93(c, ik
)-0.39 -0.60
(id, il
)0.84 0.74(
ik, b)
0.20 0.12(il/id, π
)0.40 0.30
of cross-correlations correctly. Quantitatively, though there are some variations, most of the
key business cycle relevant correlations like consumption and output, credit and output,
consumption and investment, output and inflation, and the financial cycle relevant correla-
tions like credit-to-output ratio and interest rate spread, credit-to-output ratio and inflation,
inflation and interest rate spread, co-movements of policy rate, deposit and lending interest
rates, and countercyclical movements of real variables to interest rates are explained by the
baseline parametric configuration of model.
4.1.4 Variance Decomposition Results
In Table 9, the results of the forecast error variance decomposition (FEVD) of the baseline
model are reported. Similar to the vast empirical literature on monetary transmission in
India, we find that monetary policy shock can account for only a small portion of output
fluctuations (8.72 per cent). In contrast, the shock to fiscal spending explains the variations
of aggregate output in a greater magnitude (12.27 per cent). More than 50 per cent of the
output fluctuations are explained by the technology shocks. Similar to aggregate output,
75 per cent of the variations of CPI inflation is explained by the supply side disturbances.
Monetary policy shock explains little variation of the same (5.32 per cent). Mark-up shock
and preference shock for housing appear to be negligible for the movements of real and
financial variables.
While the above set of FEVD results align with the findings of existing literature, we
obtain an interesting observation from the contribution of the shock to LTV ratio as one of the
drivers of business cycle variations. Except the policy rate, shock to LTV ratio for household
explains considerable variations across the real, nominal and financial variables of the model.
It is also modestly complemented by the similar kind of shock for the entrepreneur. This
result provides evidence for the significance of financial shocks in the Indian economy.
45
Table 9: Baseline Result of Variance Decomposition of Key Aggregates (in per cent)Variables ξA ξik ξm G ξy ξHLV ξELV ξh
y 48.95 5.58 8.72 12.27 0.25 21.59 2.64 0.00π 37.32 38.13 5.32 0.25 3.00 14.21 1.77 0.00b 21.49 24.20 20.09 0.53 0.66 29.95 3.09 0.00c 37.00 34.11 4.59 2.19 0.23 20.09 1.78 0.00l 7.62 8.59 20.26 24.32 0.58 33.16 5.46 0.00ik 22.59 57.97 3.32 0.95 0.37 10.23 4.57 0.00ih 19.07 16.26 30.16 5.83 1.03 22.07 5.58 0.00i 24.68 30.64 32.61 0.06 0.66 9.91 1.44 0.00id 28.36 32.57 4.95 12.69 1.88 12.17 7.36 0.02il 33.81 37.70 11.40 0.21 0.82 14.03 2.03 0.00
As a whole, the bulk of aggregate fluctuations is explained by the technology shocks in
the forms of total factor productivity and investment specific technology. The policy shocks
remain to be secondary drivers of the cyclical fluctuations in India. Transmission of monetary
shock is found to be substantially weak.
4.2 Transmission Mechanism of Monetary Policy Shock
In our baseline NK-DSGE model with financial frictions, monetary policy shock affects the
economy from the demand side as well as the supply side. While the demand side effect
of the policy shock works directly through the credit channel, it exerts the supply side
effect indirectly via the cost of capital and the labour market adjustments. In our model, a
positive interest rate shock by the monetary authority leads to co-movements in the credit
market interest rates, and sets in the contractionary effects on the key macroeconomic and
financial variables like credit to household and firm, consumption, investment in physical
capital and housing, labour employment, output and inflation. Although the interest rate,
asset price and expectation channels of transmission do exist in the model, the dynamics
of MPT is predominantly led by the broad credit channel. Such predominance of credit
channel of MPT is a consequence of the bank lending and the balance sheet channels. The
in-built feedback mechanism between real and financial sectors of the model pins down the
transmission process of these channels. The bank lending channel comes into action as soon
as the set of interest rates starts responding to change in the policy rate and affects the
demand for credit in the economy. Besides, the balance sheet channels become operational
primarily from the borrower’s side (household and firm) and then from the lender’s side
(bank) too. Given the borrower’s balance sheet constraint, contraction of credit demand
46
impacts the demand for consumption and investment goods, and factors of production in
the real side. By a cascading effect, contraction of credit demand also impacts the bank’s
profitability and net worth position due to presence of its balance sheet constraint. We
spell out the details of transmission process and sectoral adjustments using the IRF plots of
Figures 10 to 12 based on a positive monetary policy shock in our model.
5 10 15 200.05
0
0.05i
5 10 15 200.01
0
0.01is
5 10 15 200.02
0
0.02ib
5 10 15 200.01
0
0.01id
5 10 15 200.02
0
0.02ibh
5 10 15 200.02
0
0.02ibe
5 10 15 200.02
0
0.02il
5 10 15 200.5
0
0.5bh
5 10 15 200.02
0.01
0be
Figure 10: Effects of Monetary Policy Shock
5 10 15 200.2
0.1
0bib
5 10 15 200.2
0
0.2b
5 10 15 200.4
0.2
0pi_b
5 10 15 200.01
0
0.01c
5 10 15 200.05
0
0.05ik
5 10 15 200.1
0
0.1ih
5 10 15 20
10 3
5
0
5k
5 10 15 20
10 3
4
2
0h
5 10 15 200.02
0
0.02qh
Figure 11: Effects of Monetary Policy Shock
47
5 10 15 20
10 3
5
0
5qk
5 10 15 200.02
0
0.02u
5 10 15 200.01
0
0.01lp
5 10 15 200.05
0
0.05lip
5 10 15 200.05
0
0.05l
5 10 15 200.02
0
0.02y
5 10 15 200.01
0
0.01rw
5 10 15 200.01
0
0.01m c
5 10 15 20
10 3
5
0
5pi
Figure 12: Effects of Monetary Policy Shock
4.2.1 Transmission to Banking Sector
Policy interest rate rises and gears up the entire spectrum of interest rates of the banking
sector as an immediate effect of a monetary policy shock. At the outset, it raises the lending
and deposit rates of the wholesale branch of the bank subject to SLR, CRR and bank capital
adequacy norms. Then, it gradually passes through the deposit and lending rates of the retail
branch. The pass-through of the shock remains incomplete at the retail level interest rates
as the wholesale and retail branches face different types and degrees of financial frictions in
the form of quadratic adjustment costs. Since the interest rate adjustment cost is higher
for retail lending to the entrepreneur compared to the household, impact response of the
lending rate on firm’s borrowing is relatively lower than the household’s borrowing interest
rate subject to the respective elasticities of credit demand. Subsequently, the demand for
credit by the impatient household declines sharply as compared to the credit demanded by
the firm. Parallel to this, interest rate on deposits rises but in a modest way due to presence
of statutory norms, reserve requirements and interest rate adjustment cost.
On the whole, frictions in the bank-based credit market leads to sluggish upward move-
ments of the market interest rates and shifts the demand schedules for deposit and loan
contracts. Due to elastic nature of market demand, rising loan interest rate reduces the
bank’s earning from credit, affects its profit adversely and thus, drives down the net worth.
As apparent from the impulse response plots, interest rates on deposits show faster mean
reversion compared to the interest rates on loan contracts. Credit to household and firm
shrinks substantially, interbank borrowing by the bank decreases and banks’profitability
48
goes through a deep negative swing. Following these banking sector adjustments, real seg-
ment of our model economy starts responding to the monetary policy shock.
4.2.2 Transmission to Real Sector
Response from Demand Side: Demand side channel operates via the standard con-
sumption Euler relations, optimal conditions for the capital investment and the investment
in housing goods. As the retail deposit rate and borrowing rates rise, the opportunity cost
of current consumption and housing accumulation increase. Further, with the rising retail
borrowing rate, the borrowing constraint becomes tighter for the impatient household and
reduces their access to loanable fund. So, both the savers and borrowing households will cut
down their demand for final goods consumption and investment in housing. Entrepreneurial
consumption also follows the similar pattern. Hence, we observe a negative impact effect
on aggregate consumption. However, it is reversed in the subsequent periods and dies down
later as the impatient household enjoys a positive wealth effect for a while due to their rising
real wage from the labour market adjustments.17
In contrast to the declining demand for consumption and investment in housing, demand
for capital investment from the wholesale firms does not fall at the impact of the shock as
it is fixed by the last period’s choice of capital. For this reason, there is no change in the
stock of physical capital at the period of impact. Nevertheless, in the forthcoming quarters,
investment in capital goods starts decelerating due to increased lending rate for firm’s bor-
rowing and tightening of their collateral constraint. Therefore, impulse of the interest rate
shock goes through the contraction of credit demand from the borrowing household and firm
to contraction of the aggregate demand in the economy.
The real price of investment in housing drops following the decline in the demand for
housing accumulation as an impact effect. However, the same for physical capital remains
positive and exhibits a sharp rise at the impact due to positive investment demand. It is
pacified in the later periods with the downturn in capital investment.
Response from Supply Side: With rising loan interest rate for firms, the wholesale
entrepreneur curtails the purchase of new capital from the capital goods producer as their
demand for capital is fully backed by the borrowing from the bank subject to the periodical
LTV ratio. Since, the demand for capital is optimally set one period ahead, at the impact
of policy shock physical capital does not show any movement and stay close to zero. But
in the subsequent periods, the entrepreneur reduces the demand for borrowing subject to
17Markovic (2006) documented similar type of IRF pattern for consumption while investigating on thebank capital channel of transmission for the UK economy.
49
the collateral constraint, and we observe a steady decline in the capital goods production.
Similar to capital goods production, housing goods producing sector also faces significant
contraction with a large swing. As the cost of capital rises, demand for capital and its
utilisation descend in the intermediate goods production. Parallel to this, employment level
falls as a general equilibrium response to the contraction of derived demand for labour in
production and it drives down the aggregate output. It can be observed from the IRF plots
that impulse response of aggregate output reflects similar pattern of the impulse response of
labour employment.
4.2.3 Transmission to Inflation
In order to understand the MPT to inflation, we have to consider the internal adjustment
process in the competitive labour market as it influences the responsiveness of real wage and
real marginal cost. The role of expectation formation also comes in, which together with the
real marginal cost determines the response of inflation.
In the set-up of competitive labour market, slack in employment affects the impatient
household more adversely than the patient ones due to higher share of the first group in
the production process.18 Since the labour market is heavily populated by the borrowing
households, cut down in their employment pushes up their real wage significantly high which
raises the average real wage of the economy. Consequently, the real marginal cost of produc-
tion rises. So, we observe the IRF plot of real marginal cost to mimic the pattern of average
economy-wide real wage.19
However, this acceleration in real marginal cost does not translate into inflation as the
standard new Keynesian forward-looking expectation channel comes into action. Given the
calibrated parameters, forward-looking component occupies dominant share (nearly 65 per
cent) in expectation formation in the price-setting behaviour of the final goods producing
firm. Therefore, the price-setting retail firm adjusts its inflation expectation downward,
which takes over the momentum of real marginal cost and brings down inflation. This
adjustment in inflation dynamics takes a couple of quarters and generates a feeble positive
response at the impact level. Nevertheless, the positive effect dies down quickly and inflation
starts to decline sharply.
18National Accounts Statistics (NAS) of Ministry of Statistics and Programme Implementation (MOSPI)on household savings suggests that the proportion of households with financial saving is 23 per cent on anaverage over the period of 2011-12 to 2015-16. This estimate lies in the confidence interval of our modelestimated result of saver’s proportion (i.e., 21 to 45 per cent). This observation closely supports the factthat proportion of saver is much lesser than the proportion of borrower in the economy.19In a cross-country study, Normandin (2006) observed positive response of real wage with respect to a
positive interest rate shock. He found this result to be consistent with the models of nominal and financialfrictions.
50
To summarise the MPT results of our model, a positive interest rate shock leads to output
contraction and restrains inflation by raising the market interest rates, squeezing the supply
of credit, and shrinking the derived demand for factors of production.
4.3 Financial Frictions andMonetary Transmission: Evidence from
Counterfactual Experiments
Our baseline model is characterised by the real, nominal and financial frictions which can po-
tentially determine the pass-through of a monetary policy shock. Given the objective of this
study, we focus on the role of frictions that are directly or indirectly related to the financial
sector. Since the commercial banking led credit market depicts the financial sector in our
model, we examine different forms of frictions associated with the credit market activities.
These frictions are pertaining to either the price of financial resources or availability of the
same. Nevertheless, they have distinguishing implications for the transmission mechanism
of a monetary policy shock. As these frictions are captured by a set of parameters, we con-
duct few counterfactual experiments with respect to baseline values of those parameters and
investigate their resultant effects on the transmission mechanism.
We examine the change of accumulated effects of a positive monetary policy shock for
different counterfactual experiments and draw our inference accordingly. The accumulated
effects of monetary policy shock are taken over the period of eight quarters. In the sensitivity
experiment, we reduce the friction parameters one at a time and document the corresponding
accumulated effects of monetary policy shock on output, inflation and credit. By comparing
these new accumulated effects with the baseline ones, as presented in Table 10, we identify
the changes in the magnitude of monetary transmission. In order to understand the policy
relevance of these financial market frictions more comprehensively, we evaluate the elasticity
of monetary transmission in terms of accumulated effects on output and inflation with respect
to each parameter and reported in Table 11. This exercise provides a quantitative assessment
for the respective role of different friction components for the transmission mechanism in
India.
We start with the set of friction parameters that are related to price-setting actions of the
financial products like deposit and loan contracts. In our model, there are three adjustment
cost parameters associated with the optimal choice of interest rates on the deposits of patient
household (φd) and loans for the borrowing household (φbh) and firm (φbe). In case of interest
rate adjustment cost for deposits, the result suggests that the frictionless state (φd = 0) has
51
Table 10: Sensitivity Experiments on Accumulated Effects of Monetary Policy Shock (in percent)
Variables y π bBaseline -6.78 -0.88 -67.91φd= 0 -6.80 -0.87 -67.88φbh= 0 -6.73 -0.91 -68.89φbe= 0 -6.47 -1.06 -68.66φz= 0 -11.37 0.49 -107.07
ξhLV = 0.65 -4.28 -1.22 -64.21γR= 0.10 -8.44 -0.92 -74.06γ = 0.83 -1.20 -1.13 -44.78
negligible impact on the pass-through of policy shock. This may be attributed to the low
base of depositors, which is found to be 19.8 per cent of the labour market population.20
However, in case of the lending rates for household and firm, reductions in interest rate
adjustment cost improve the transmission process except for credit. From Table 10, it is
apparent that contractionary effect of a positive interest rate shock becomes moderated for
output and turns out to be more intensive for inflation in absence of the adjustment costs.
In absence of the adjustment costs, flexible retail lending rates lead to faster mobilisation of
financial resources and better allocation of the factors of production, and generates stronger
effect on inflation reduction with weaker effect for output contraction.
Next, we examine the frictions related to availability of credit. There are two items
in the checklist: one is adjustment cost parameter related to maintaining of bank capital
adequacy norm and the other is steady-state LTV ratio for collateral constraints. We do the
sensitivity experiments for (i) zero adjustment cost for maintaining the bank capital adequacy
requirement (φz = 0) and (ii) relaxing the collateral constraint for borrowing household by
raising the steady-state LTV(ξhLV = 0.65
).
From the change of accumulated effects, it can be noticed that the transmission of mone-
tary policy shock becomes more pronounced for both cases though in different directions and
different magnitudes. In absence of adjustment cost for maintaining bank capital adequacy
requirement, contractionary effect on output and credit deepen substantially compared to
the other types of friction components. Besides, it creates inflationary pressure to some
extent. This counter-intuitive result appears due to presence of the large segment of credit-
constrained borrowers. In absence of the adjustment cost for restoring capital adequacy
20To cross-examine this counterfactual result, we look at the accumulated effects of a monetary policyshock for φd = 0 with γ = 0.83. We find that credit (-44.78 per cent) and output (-1.2 per cent) fallsubstantially lesser while the inflation declines in a large extent (-1.12 per cent) as compared to the baselineresults.
52
requirement, net worth of the bank improves, which subsequently strengthens the bank cap-
ital channel of monetary transmission. However, such improvement of transmission process
is nullified due to presence of the large section of borrowing households in the labour market.
The predominance of borrowing households augments the contraction of aggregate demand
via demand for credit and leads to pervasive response of real wage in the labour market.
In combination of these two actions simultaneously, the improvement of transmission under
frictionless state for bank capital adequacy does not show up in the counterfactual results.21
The easing of collateral constraint for borrowing household clearly produces a favourable
impact for the monetary transmission to output and credit with a significant fall in inflation.
The reason is relatively straight forward. Higher long-run LTV provides borrowing household
greater access to credit and strengthens the credit channel of transmission. This improves
the transmission process of policy shock in the economy.
Looking into the household structure of the underlying economy, it is notable that pres-
ence of liquidity-constrained household and large proportion of borrowing household in the
labour market create serious bottleneck for the pass-through of monetary policy shock. Typ-
ically, the proportions of liquidity-constrained household (γR = 0.40) and impatient / bor-
rowing household (γ (1− γR) = 0.402) are inversely related to the MPT mechanism.22 With
declining share of liquidity-constrained household, the transmission of monetary policy im-
proves in the economy. This result supports the fact that greater financial inclusion leads
to better transmission process of the policy shock. Along with the proportion of liquidity-
constrained household, the share of patient household or saver (19.8 per cent) vis-à-vis
impatient household plays an important role for the monetary transmission. As it is evi-
dent from our estimation results, the proportion of borrowers is higher than savers in the
Indian economy compared to the other economies. This has two implications. First, it limits
the scope of the bank to mobilise deposits for loanable fund. Second, there exists relatively
larger proportion of credit-constrained borrowers in the competitive labour market, who play
a critical role for determining the response of the overall employment, real wage (thereby,
real marginal cost and inflation) and finally, aggregate output with respect to a monetary
policy shock. Given that a positive policy shock is in place, the contractionary impact on
aggregate demand subsequently leads to reduction in the derived demand for labour employ-
ment in the labour market. Since, the credit-constrained borrowers occupy the greater share
21For the validation of our argument regarding this counterfactual result, we look at the accumulated effectsof a monetary policy shock for φz = 0 with γ = 0.83. Although the inflationary response in transmissiondoes not revert to be negative (0.22 per cent), we find that transmission to credit (-58.32 per cent) andoutput (-2.79 per cent) improves.22Note that low value of γ indicates greater proportion of borrowers compared to savers in the population
and vice versa.
53
in the production process, they face the fierce hit of the employment cut, which is pivotal
for output contraction. Hence, in our counterfactual experiment with γ = 0.83, we observe
little contraction of credit and output and significant reduction of inflation.
Table 11: Financial Frictions and Responsiveness of Monetary Policy TransmissionFriction Elasticity of Elasticity of
Parameters Output Effect Inflationary Effectφd 0 0.01φbh 0.01 -0.03φbe 0.05 -0.21φz -0.68 1.56ξhLV -2.03 2.09γR 1.15 -0.05γ -0.54 0.18
To summarise our observations on the role of different frictions related to financial activ-
ities, we compute the elasticity of monetary transmission to output and inflation based on
the accumulated effect of the shock.23 The tabulated numbers in Table 11 reveal how the
degree of contractionary effect of a monetary policy shock alters with respect to one unit
change in the value of respective friction parameter. The value of elasticity also suggests the
quantitative importance of the corresponding friction. It is noticeable that except the fric-
tions for deposit rate adjustment and bank capital adjustment, elimination of the financial
frictions from the economy reduces the contractionary effects of monetary policy shock on
the real output and stabilises inflation to a larger extent. Given the size of the elasticities,
it appears that collateral constraint, financially excluded population and low base of depos-
itors play a major role in hindering the transmission process and need more attention from
the policymakers. Frictions related to price of financial resources, albeit quantitatively less
significant, play a moderate role to improve the pass-through of the transmission mechanism.
4.4 Policy Implications
In the preceding subsections, we have illustrated the mechanism of MPT and examined the
role of different forms of financial frictions. In light of our findings, one would be curious23In Appendix E, the IRF plots of our counterfactual experiments are reported. Blue lines denote low
level of friction and red lines denote the high degree of friction. In Figures 13 to 16, we have presented thechanges of propagation mechanism of monetary policy shock with respect to changes of four major frictionparameters of our analysis: ξhLV , γR, γ and φz. In Figure 17, we have presented a case of bank capital channelof MPT due to change in the macroprudential policy-driven financial friction of capital adequacy norm (κb).The IRF plots reveal that the presence of CAR works as a friction in the MPT mechanism.
54
to envisage the policy implications of the same. To this end, we provide a brief discus-
sion on the policy relevance of our study. First, policy intervention is required to bring
financially excluded households into the coverage of the banking system. In this regard,
the Jan Dhan Yojana of the government and other policy initiatives of the RBI in recent
years are likely to imrpove the policy transmission mechanism in the coming years. Second,
the results suggest that a reduction in adjustment costs related to bank lending activities
along with an increase in the depositors’base can lead to a better tranmission of mone-
tary policy. This can be achieved by improving the levels of operational and managerial
effi ciency in mobilisation and allocation of the financial resources as well as by protecting
the depositors’ interest. Typically, the adjustment cost related to lending is high partly
due to operational/menu costs, processing fees, documentation charges, stamp duty and so
on. The use of data analytics may help in automation, which can successively reduce the
adjustment costs and hence, the degree of frictions. Third, our results also suggest that
collateral constraint is one of the major sources of financial frictions that lead to incomplete
pass-through of policy actions. The removal of collateral constraints by increasing the LTV
for households and firms substantially improves the MPT. However, an increase in LTV may
have some adverse implications for the financial stability. Hence, there is a trade-off between
monetary policy and macro-prudential policy measures. The trade-off can partly be taken
care through designing the policies to improve transparency and disclosures, enforcement of
laws, market discipline and strong corporate governance by using the credit registries, social
and/or venture networks with better enforcement and information tools and social norms so
that the rise in LTV does not increase risks to the financial system. Besides, the use of big
data may help in getting information about the borrowers related to their income, spending,
business operations and other behavioural details. This would enable the lenders to predict
the repayment behaviour of borrower, take more informed and profitable credit decisions in
real time, and set the collateral requirement for the prospective borrowers more objectively.
Apart from the above mentioned policy implications, our model can be instrumental to
examine a variety of policy rules in order to ensure the best possible outcome for macroeco-
nomic stabilisation. With reference to the baseline model, we explore the alternative forms
of monetary policy rules by augmenting the Taylor rule specification with financial variables.
In Table 12, we list the different forms of asset price and credit augmented monetary policy
rules, which are experimented with the baseline model one at a time. The motivation is to
check whether the financial variable augmented policy rule can produce a better outcome in
terms of economic stabilisation given the choice of policy frameworks. For this purpose, we
consider the central bank loss function (LC) based on variances of inflation and output as
given below:
55
LC = σ2π + αwσ2y (74)
where, αw > 0 determines the choice of policy framework according to the relative weigh-
tage attached with the policy objectives. When 0 < αw < 1; it implies that the policy
authority is inclined to minimise its welfare loss incurred due to variability of inflation more
rather than the variability of output and hence, attaching higher weightage to inflation sta-
bilisation relative to output stabilisation (IT framework). The situation will be reversed
when αw > 1, which implies that output stabilisation is relatively more desirable to the
policy authority instead of stabilising inflation (YT framework). Finally, for αw = 1, the
policy authority remains indifferent or in a neutral position with respect to stabilising its
policy objectives (Neutral framework).
Table 12: Monetary Policy Rules Augmented by Financial VariablesPolicies Taylor Rule Specifications Parameters
Rule 1 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕyYt
]+εm,t Baseline V alues
Rule 2 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕyYt + ϕfQ
ht
]+εm,t ϕf = 1.15
Rule 3 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕyYt + ϕfQ
kt
]+εm,t ϕf = 1.15
Rule 4 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕyYt + ϕf Bt
]+εm,t ϕf = 1.15
Rule 5 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕy
(Bt − Yt
)]+εm,t ϕy = 0.25
Rule 6 ıt= ϕiıt−1+ (1− ϕi)[ϕπEt {πt+1}+ ϕyYt + ϕf
(Bt − Yt
)]+εm,t ϕy = 0.25; ϕf = 0.1
For the policy experiment, first, we set up three cases: αw = 0.5, αw = 1 and αw = 1.5
denoting three different frameworks in terms of policy objectives. Then, we simulate the
baseline model with policy rules 1 to 6 one at a time, record the model generated volatilities
of inflation and output, and compute the hypothetical welfare loss of the central bank subject
to the choice of αw. Rule 1 is the baseline policy rule of standard form with interest rate
smoothing, inflation and output. Following Table 4, we use the estimates of policy coeffi cients
(i.e., ϕi, ϕπ, and ϕy). In Rules 2 and 3, we extend the baseline policy rule by adding the
asset prices, i.e., real prices of housing (Qht ) and physical capital (Q
kt ), respectively. In Rule
4, we consider the case when policy rate responds to the movements of credit cycle (Bt).
Rule 5 depicts the scenario when the central bank responds to the deviation of credit-to-
output ratio from its steady-state level instead of output only. Modifying Rule 5 with output
targeting component along with the credit-to-output ratio, we present Rule 6 as suggested
56
by Badarau and Popescu (2012). Following Castelnuovo (2013), we calibrate the value of ϕfas 1.15 except for Rule 6. In case of Rule 6, we set ϕf at 0.1 as suggested in Badarau and
Popescu (2014).
The results of model simulation with alternative monetary policy rules are presented
in Table 13. Except the case of housing price augmented Taylor rule, it is found that
the standard form of Taylor rule with forecast-based inflation and contemporaneous output
stands out as the optimal one across all the policy frameworks. Housing price augmented
Taylor rule (i.e., Rule 2) performs marginally better than the conventional Taylor rule as it
reduces the volatility of output mildly. In contrast, adjusting policy interest rate to smoothen
the credit cycle (Rules 4 to 6) does not seem to be useful. In fact, it exacerbates the volatilities
of inflation and output. Moreover, comparing three different policy frameworks we find that
inflation stabilisation is the most desirable policy option for the central bank as it leads
to minimum welfare loss irrespective of the policy rules. Overall, it appears that targeting
financial variables in the monetary policy rule may not be appropriate for the purpose of
economic stabilisation.
Table 13: Alternative Policy Rules and Central Bank Welfare LossDifferent σ2π σ2y CB Loss (in %)
Policy Rules (in %) (in %) IT (αw = 0.5) Neutral (αw = 1) YT (αw = 1.5)1 0.39 1.10 0.94 1.49 2.042 0.39 1.08 0.93 1.47 2.013 0.60 1.17 1.19 1.77 2.364 1.87 1.39 2.57 3.26 3.965 0.53 1.24 1.15 1.77 2.396 0.56 1.12 1.12 1.68 2.24
5 Conclusion
Recently, the conduct of monetary policy in India has undergone a fundamental institutional
reform. The flexible inflation targeting (FIT) has been formally adopted by the RBI. Subse-
quently, a monetary policy committee (MPC) has been constituted to determine the policy
rate in order to achieve the inflation target. Under FIT framework, the RBI is assigned to
achieve an inflation target of 4 per cent with a band of +/- 2 per cent. Any failure to achieve
the same for three consecutive quarters needs to be explained in the Parliament with the
reasons for the failure. In addition, the remedial measures and the expected time line to
57
return to the targeted inflation rate have to be proposed. Such institutional mandate clearly
reflects the need for faster and effi cient transmission mechanism in the economy. In this
regard, understanding the frictions embedded at the different layers of bank-led credit mar-
ket in the financial sector can contribute towards better decision making for the monetary
authority. With this motivation, we studied the role of different structural bottlenecks and
institutional impediments of the credit market arrangements in the transmission channels of
the monetary policy.
Examining the friction of financial sector is important as this type of friction debilitates
the pass-through of monetary policy both from the policy instrument to intermediate targets
and intermediate targets to the policy objectives. In this study, we have addressed this issue
using an NK-DSGE model with the banking sector. The model is augmented with the
Indian economy-specific features, estimated with the quarterly data (1999:Q4 to 2015:Q3)
and validated with the business cycle facts of the key macroeconomic and financial variables.
The baseline model replicates a set of stylised facts like: (i) co-movements of the interest
rates with incomplete pass-through, (ii) countercyclical movement of interest rate spread,
(iii) presence of the broad credit channel in combination with interest rate channel, and (iv)
weak pass-through of MPT to output and inflation. In addition, it identifies a pervasive
response of real wage in the labour market adjustment and underscores the role of forward-
looking expectation for stabilising inflation. Considering the degree of MPT in terms of
accumulated effects obtained from the ‘baseline model’as the benchmark, we undertake the
counterfactual experiments on the elements of financial frictions. Our experiment provides
a comparative analysis on the different credit market frictions which helps evaluating their
respective role and quantitative significance. It is found that the collateral constraints and
financially excluded segment of the economy cause major obstacles in the MPT while interest
rate rigidity on the lending rates is of secondary importance.
There are a few caveats to be mentioned regarding our study. In the modelling frame-
work, structure and implications of the interbank market and soveriegn bond market are not
addressed explicitly. Oligopolistic competition of the banking sector and increasing role of
the non-bank financial corporations are missing, which would provide more appropriate de-
piction of the Indian financial sector instead of the monopolistically competitive commercial
banking sector. Also, the model does not take into account any labour market frictions. In
the side of quantitative analysis, the baseline parameterisation of the model is done using
Bayesian methodology, which is sensitive to the choice of prior distributions and the his-
torical data series of the observables. Hence, results of the study critically depend on the
microfoundation and parametric configuration of the model.
The study can be enriched if the asymmetric pass-through of the monetary policy shock
58
to the retail interest rates of the commercial bank, subject to the market liquidity, is ex-
plained. Besides, the role of informal credit market can be examined as it is quite relevant
for India. One can also extend the baseline model to the open economy set-up. All these
potential extensions may improve the fitness of the model and provide useful insights for the
transmission mechanism of monetary policy in the economy.
59
6 Bibliography
Acharya, V., (2017) Monetary transmission in India: Why is it important and why hasn’t
it worked well? Inaugural Aveek Guha Memorial Lecture, Homi Bhabha Auditorium, Tata
Institute of Fundamental Research, Mumbai, November.
Agénor, P. R. & Pereira da Silva, L. A. (2014) Macroprudential regulation and the
monetary transmission mechanism. Journal of Financial Stability, 13 (C), pp. 44-63.
Al-Mashat, R. (2003) Monetary policy transmission in India: Selected issues and statis-
tical appendix. Country report no. 03/261, International Monetary Fund.
Aleem, A. (2010) Transmission mechanism of monetary policy in India. Journal of Asian
Economics, 21(2), pp. 186-197.
Altunbas, Y., Gambacorta, L., & Marques-Ibanez, D., (2009) Securitisation and the bank
lending channel. European Economic Review, 53(8), pp. 996-1009.
Anand, R., Delloro, V., & Peiris, S. J., (2014) A credit and banking model for emerg-
ing markets and an application to the Philippines, in The Evolving Role and Limits of
Monetary Policy: New Perspectives for Emerging Market Economies, 2014 BSP Interna-
tional Research Conference Proceedings, Bangko Sentral ng Pilipinas International Research
Conference, Manila, Philippines, published in Issue No. 2, September 2015, pp. 191-230.
Available at: http://www.bsp.gov.ph/downloads/Publications/2014/IRC2014.pdf
Anand, R., Peiris, S., & Saxegaard, M., (2010) An estimated model with macrofinancial
linkages for India. IMF Working Paper No. WP/10/21, International Monetary Fund.
Andrés, J., & Arce, O. (2012). Banking competition, housing prices and macroeconomic
stability. The Economic Journal, 122 (565), pp. 1346-1372.
Aysun, U., Brady, R., & Honigc, A., (2013) Financial frictions and the strength of mon-
etary transmission. Journal of International Money and Finance, 32 (February), pp. 1097-
1119.
Badarau, C., & Popescu, A., (2014) Monetary policy and credit cycles: A DSGE analysis.
Economic Modelling, 42, pp. 301-312, October.
Banerjee, S. and Basu, P., (2017) Technology shocks and Business cycles in India. Macro-
economic Dynamics, pp. 1-36. doi:10.1017/S1365100517000438
Bernanke, B. S., (1993) How important is the credit channel in the transmission of
monetary policy? : A comment. Carnegie-Rochester Conference Series on Public Policy, 39
(1), pp. 47-52, December.
Bernanke, B., & Blinder, A. S., (1992) The federal funds rate and the channels of mone-
tary transmission. American Economic Review, 82(4), pp. 901-921.
Bernanke, B., & Gertler, M., (1995) Inside the black box: The credit channel of monetary
60
policy transmission. Journal of Economic Perspectives, 9 (4), pp. 27-48.
Bernanke, B., & Gertler, M., (1989) Agency costs, net worth, and business fluctuations.
American Economic Review, 79(1), pp. 14-31.
Bernanke, B. S., Gertler, M., & Gilchrist, S., (1999) The financial accelerator in a quan-
titative business cycle framework. In Handbook of Macroeconomics, (Eds.) John Taylor and
Michael Woodford, pp. 1341-93, Amsterdam: North Holland.
Bhattacharya, R., Patnaik, I., & Shah, A., (2010) Monetary policy transmission in an
emerging market setting. IMF Working Paper, No: WP/11/5.
Bhattacharya, R., & Patnaik, I., (2015) Financial inclusion, productivity shocks, and
consumption volatility in emerging economies. World Bank Economic Review, 30 (1), pp.
171-201.
Bhattarai, S., Lee, J. W., & Park, W. Y. (2015) Optimal monetary policy in a currency
union with interest rate spreads. Journal of International Economics, 96(2), pp. 375-397.
Bhaumik, S. K., Dang, V., & Kutan, A. M., (2011) Implications of bank ownership for
the credit channel of monetary policy transmission: evidence from India. Journal of Banking
and Finance, 35 (9), pp. 2418-28.
Blinder, A. S., (2000) Central-bank credibility: Why do we care? how do we build it?.
American Economic Review, 90 (5), pp. 1421-1431, December.
Blum, J., & Hellwig, M., (1995) The macroeconomic implications of capital adequacy
requirements for banks. European Economic Review, 39 (3-4), pp. 739-749.
Borio, C., & Zhu, H., (2012) Capital regulation, risk-taking and monetary policy: A
missing link in the transmission mechanism?. Journal of Financial Stability, 8 (4), pp.
236-251.
Brooks, S. P., & Gelman, A., (1998) General methods for monitoring convergence of
iterative simulations. Journal of Computational and Graphical Statistics, 7 (4), pp. 434-455.
Brunnermeier, M. K., Eisenbach, T. M., & Sannikov, Y., (2012) Macroeconomics with
financial frictions: A survey. NBER Working Papers 18102, National Bureau of Economic
Research, Inc.
Brzoza-Brzezina, M., & Kolasa, M. (2013) Bayesian evaluation of DSGE models with
financial frictions. Journal of Money, Credit and Banking, 45(8), pp. 1451-1476.
Canzoneri, M., Collard, F., Dellas, H., & Diba, B. (2016) Fiscal multipliers in recessions.
The Economic Journal, 126 (590), pp. 75-108.
Carlstrom, C. T., & Fuerst, T. S., (1997) Agency costs, net worth, and business fluctua-
tions: A computable general equilibrium analysis. American Economic Review, 87 (5), pp.
893-910.
Cecchetti, S. G., (1995) Distinguishing theories of the monetary transmission mechanism.
61
Federal Reserve Bank of St. Louis Economic Review, pp. 83-100.
Cecchetti, S. G., & Krause, S., (2001) Financial structure, macroeconomic stability, and
monetary policy. NBER Working Paper No. 8354.
Christiano, L. J., Eichenbaum, M., & Evans, C. L., (1999) Monetary policy shocks: What
have we learned and to what end? In Handbook of Macroeconomics, 1 (A), (Eds.) John B.
Taylor and Michael Woodford, Amsterdam.
Chui, M., Kuruc, E., & Turner, P., (2016) A new dimension to currency mismatches
in the emerging markets - non-financial companies. BIS Working Papers 550, Bank for
International Settlements.
Claessens, S., Demirgüç-Kunt, A., & Huizinga, H. (2001) How does foreign entry affect
domestic banking markets? Journal of Banking & Finance, 25 (5), pp. 891-911.
Cooley, T., Marimon, R., & Quadrini, V., (2004) Aggregate consequences of limited
contract enforceability. Journal of Political Economy, 112 (4), pp. 817-847.
Copaciu, M., Nalban, V., & Bulete, C., (2015) An estimated DSGE model for Romania.
Dynare Working Papers Series No. 48, November.
Cúrdia, V., & Woodford, M. (2016). Credit frictions and optimal monetary policy. Jour-
nal of Monetary Economics, 84, pp. 30-65.
Cúrdia, V., & Woodford, M., (2010) Credit spreads and monetary policy. Journal of
Money, Credit and Banking, 42 (1), pp. 3-35.
Das, S., (2015) Monetary policy in India: Transmission to bank interest rates. IMF
Working Papers 15/129, International Monetary Fund.
Das, A., Mishra, P., & Prabhala, N., (2015) The bank lending channel of monetary
transmission: New evidence from India, mimeo.
Dib, A. (2010). Capital requirement and financial frictions in banking: macroeconomic
implications. Bank of Canada Working Paper, No. 2010, 26.
Dynare User Guide: An introduction to the solution and estimation of DSGE models,
copyright by Thomas Mancini Grioli, January 2003 version.
Fazzari, S. M., Hubbard, R. G., & Bruce, C. P., (1988) Financing constraints and corpo-
rate investment. Brookings Papers on Economic Activity, Economic Studies Program, The
Brookings Institution, 19 (1), pp. 141-206.
Gabriel, V., Levine, P., Pearlman, J., & Yang, B., (2012) An estimated DSGE model of
the Indian economy. In The Oxford Handbook of the Indian Economy, (Ed.) Chetan Ghate,
pp. 835-890, Oxford University Press: New York.
Gambacorta, L., (2008) How do banks set interest rates?. European Economic Review,
52 (5), pp. 792-819.
Gambacorta, L., & Signoretti, F. M., (2014) Should monetary policy lean against the
62
wind?. Journal of Economic Dynamics and Control, 43, pp. 146-174.
Gerali, A., Neri, S., Sessa, L., & Signoretti, F., (2010) Credit and banking in a DSGE
model of the Euro area. Journal of Money, Credit and Banking, 42 (6), pp. 107-141.
Gertler, M., & Gilchrist, S., (1993) Monetary policy, business cycles and the behavior
of small manufacturing firms. Finance and Economics Discussion Series 93-4, Board of
Governors of the Federal Reserve System (U.S.).
Gertler, M., & Kiyotaki, N., (2010) Financial intermediation and credit policy in business
cycle analysis, Chapter 11. Handbook of Monetary Economics, 3, pp. 547-599.
Gertler, M., & Karadi, P., (2011) A model of unconventional monetary policy. Journal
of Monetary Economics, 58 (1), pp. 17-34.
Goodfriend, M & McCallum, B. T., (2007) Banking and interest rates in monetary policy
analysis: A quantitative exploration. Journal of Monetary Economics, 54 (5), pp. 1480-1507.
Guerrieri, L., & Iacoviello, M., (2017) Collateral constraints and macroeconomic asym-
metries. Journal of Monetary Economics, 90, October, pp.28-49.
Hall, R. E., (2011) The high sensitivity of economic activity to financial frictions. The
Economic Journal, 121 (552), pp. 351-378.
Hart, O., & Moore, J., (1994) A theory of debt based on the inalienability of human
capital. Quarterly Journal of Economics, 109 (4), pp. 841-79.
Heathcote, J., Storesletten, K., & Violante, G. L., (2009) Quantitative macroeconomics
with heterogeneous households. Annual Review of Economics, 1 (1), pp. 319-354.
Hubbard, R. G., (1995) Is there a ‘Credit Channel’for monetary policy? Federal Reserve
Bank of St. Louis Review 77 (May/June), pp. 63-77.
Iacoviello, M., (2015) Financial business cycles. Review of Economic Dynamics, 18 (1),
pp.140-163.
Iacoviello, M., & Neri, S., (2010) Housing market spillovers: Evidence from an estimated
DSGE Model. American Economic Journal: Macroeconomics, 2 (April): pp. 125-164.
Ireland, P., (2008) The monetary transmission mechanism. In The New Palgrave Dictio-
nary of Economics, Second Edition, (Eds.) Lawrence Blume and Steven Durlauf, Hampshire:
Palgrave Macmillan Ltd.
Iskrev, N., (2010a) Evaluating the strength of identication in DSGE models. An a priori
approach. Working Papers w201032, Banco de Portugal, Economics and Research Depart-
ment.
Iskrev, N., (2010b) Local identification in DSGEmodels. Journal of Monetary Economics
57, pp. 189-202.
Iskrev, N., & Ratto, M., (2010a) Computational advances in analyzing identification of
DSGE models. 6th DYNARE Conference, June 3-4, Gustavelund, Tuusula, Finland. Bank
63
of Finland, DSGE-net and Dynare Project at CEPREMAP.
Iskrev, N., and Ratto, M., (2010b) Identication toolbox for DYNARE. 1st MONFISPOL
Conference, London, 4-5 November. The London Metropolitan University and the European
Research project (FP7-SSH) MONFISPOL.
Jain, R., & Khundrakpam, J., K., (2012) Monetary policy transmission in India: A peep
inside the black box. MPRA Paper No. 50903, June. Available at: http://mpra.ub.uni-
muenchen.de/50903/
Jermann, U., & Quadrini, V., (2012) Macroeconomic effects of financial shocks. American
Economic Review, American Economic Association, 102 (1), pp. 238-271, February.
Joshi, H., Saggar, M., & Ray, P., (1998) New monetary transmission channels: Role
of interest rates and exchange rate in conduct of Indian monetary policy. Economic and
Political Weekly, 33 (44), pp. 2787-2794.
Kapur, M., & Behera, H., (2012) Monetary transmission mechanism in India: A quarterly
model. RBI Working Paper Series, W.P.S (DEPR) : 09/ 2012, Mumbai.
Khundrakpam, J. K., (2011) Credit channel of monetary transmission in India - How
effective and long is the lag?. Working Paper (DEPR) 20/2011, Reserve Bank of India,
Mumbai.
Kashyap, A. K., & Stein, J. C., (2000) What do a million observations on banks say about
the transmission of monetary policy?. American Economic Review, 90 (3), pp. 407-428.
Kashyap, A. K., & Stein, J. C., (1995) The impact of monetary policy on bank balance
sheets. Carnegie-Rochester Conference Series on Public Policy, 42 (1), pp. 151-195.
Kashyap, A. K., & Stein, J. C., (1993) Monetary policy and bank lending. NBER
Working Paper No. 4317, April.
Kashyap, A. K., & Stein, J. C., & Wilcox, D. W., (1993) Monetary policy and credit
conditions: Evidence from the composition of external finance. American Economic Review,
83 (1), pp. 78-98.
Kletzer, K., (2012) Financial frictions and monetary policy transmission in India. The
Oxford Handbook of the Indian Economy, (ed) Chetan Ghate.
DOI: 10.1093/oxfordhb/9780199734580.013.0022
Khwaja, A. I., & Mian, A., (2008) Tracing the impact of bank liquidity shocks: Evidence
from an emerging market. American Economic Review, 98 (4), pp. 1413-1442.
Kishan, R. P., & Opiela, T. P., (2000) Bank size, bank capital, and the bank lending
channel. Journal of Money, Credit and Banking, 32 (1), pp. 121-141.
Kiyotaki, N., & Moore, J. H., (2012) Liquidity, business cycles and monetary policy.
NBER Working Paper No. 17934, DOI: 10.3386/w17934
Kiyotaki, N., & Moore, J. H., (1997) Credit cycles. Journal of Political Economy, 105
64
(2), pp. 211-248.
Kiyotaki, N., (2012) A mechanism design approach to financial frictions. In: Allen F.,
Aoki M., Fitoussi JP., Kiyotaki N., Gordon R., Stiglitz J.E. (eds) The Global Macro Economy
and Finance. International Economic Association Series. Palgrave Macmillan, London
Lahiri, A. and Patel, U. (2016) The challenges of monetary policy in emerging economies,
in Monetary Policy in India: A Modern Perspective, Springer Verlag: India, December.
Mandelman, F. S., (2010) Business cycles and monetary regimes in emerging economies:
A role for a monopolistic banking sector. Journal of International Economics, 81 (1), pp.
122-138, May.
Markovic, B., (2006) Bank capital channels in the monetary transmission mechanism.
Bank of England Working Paper No. 313, November.
Mateju, J. (2013) Explaining the strength and the effi ciency of monetary policy trans-
mission: A panel of impulse responses from a time-varying parameter model. IES Working
Paper, (No. 18/2013).
Mendoza, E., (2010) Sudden stops, financial crises, and leverage. American Economic
Review, American Economic Association, 100 (5), pp. 1941-1966, December.
Merola, R. (2015) The role of financial frictions during the crisis: An estimated DSGE
model. Economic Modelling, 48, pp. 70-82.
Mian, A., Sufi, A. & E. Verner (2017) Household debt and business cycles worldwide.
The Quarterly Journal of Economics, 132 (4), pp. 1755—1817.
Mishra, P., Montiel, P., & Sengupta, R., (2016) Monetary transmission in developing
countries: Evidence from India, In Monetary Policy in India: A Modern Macroeconomic
Approach (Eds.) Chetan Ghate and Ken Kletzer, pages 59-110, Springer Verlag: New Delhi.
Mishra, P., & Montiel, P., (2013) How effective is monetary transmission in developing
countries? A survey of the empirical evidence. Economic Systems, 37 (2), pp. 187-216.
Mishkin, F. S., (1996) Channels of monetary transmission and lessons for monetary policy.
NBER Working Paper No.5464, National Bureau of Economic Research.
Mohan, R., 2006 Evolution of central banking in India, based on lecture delivered at a
seminar organized by the London School of Economics and the National Institute of Bank
Management at Mumbai, India, on January 24.
Mohanty, M., & Kumar, R., (2016) Financial intermediation and monetary policy trans-
mission in EMEs: What has changed since the 2008 crisis ?. In Monetary Policy in India:
A Modern Macroeconomic Approach (Eds.) Chetan Ghate and Ken Kletzer, pages 111-150,
Springer Verlag: New Delhi.
Mohanty, M.S., & Turner, P., (2008) Monetary policy transmission in emerging market
economies: What is new?, in Transmission Mechanisms For Monetary Policy In Emerging
65
Market Economies, BIS Papers No. 35, Bank for International Settlements.
Mohanty, D., (2012) Evidence of interest rate channel of monetary policy transmission
in India. RBI Working Paper Series No. 06
Normandin, M., 2006. The Effects of Monetary-Policy Shocks on Real Wages: A Multi-
Country Investigation. Available from: https://core.ac.uk/download/pdf/6689285.pdf
Pandit, B. L., & Vashisht, P., (2011) Monetary policy and credit demand in India and
some other EMEs. ICRIER Working Paper No. 256, ICRIER, New Delhi.
Pandit, B.L., Mittal, A., Roy, M., & Ghosh, S., (2006) Transmission of monetary policy
and the bank lending channel: Analysis and evidence for India. Development Research Group
Study No. 25. Department of Economic Analysis and Policy Reserve Bank of India Mumbai
(January).
Patra, M., & Kapur, M., (2010) A monetary policy model without money for India. IMF
Working Papers, Working Paper No. 10/183.
Available at SSRN: https://ssrn.com/abstract=1669864
Prasad, A. & Ghosh, S., (2005) Monetary policy and corporate behavior in India, IMF
Working Paper No. 25, Washington DC.
Quadrini, V., (2011) Financial frictions in macroeconomic fluctuations. Economic Quar-
terly, 97 (3), pp. 209—254.
Rajan, R. G., (2005) Has financial development made the world riskier?. Proceedings
- Economic Policy Symposium - Jackson Hole, Federal Reserve Bank of Kansas City, issue
August, pp. 313-369.
Rotemberg, J., (1982) Monopolistic price adjustment and aggregate output. Review of
Economic Studies, Vol. 49(4), pp. 517-531.
Sahu, J. P., (2013). Inflation dynamics in India: A hybrid New Keynesian Phillips Curve
approach. Economics Bulletin, 33 (4), pp. 2634-2647.
Sen Gupta, A., & Sengupta, R., (2014) Is India ready for flexible inflation-targeting?
MPRA Paper 56430, University Library of Munich, Germany.
Available at: https://mpra.ub.uni-muenchen.de/56430/1/MPRA_paper_56430.pdf
Silva, M. da., Andrade, J., Silva, G., Brandi, V., (2012) Financial frictions in the Brazilian
banking system: A DSGE model with Bayesian estimation, Annals of Brazilian Econometric
Society Meeting.
Singh, B., (2011) How asymmetric is the monetary policy transmission to financial mar-
kets in India? Reserve Bank of India Occasional Papers, 32 (2), Monsoon.
Singh, K., & Kalirajan, K., (2007) Monetary transmission in post-reform India: an eval-
uation. Journal of the Asia Pacific Economy, 12, 158—187.
Sousa, J. M., & Zaghini, A., (2007) Global monetary policy shocks in the G5: A SVAR
66
approach. Journal of International Financial Markets, Institutions and Money, 17 (5), pp.
403-419.
Stiglitz, J.E., &Weiss, A., (1981) Credit rationing in markets with imperfect information.
American Economic Review, 71 (3), pp. 393-410.
Taylor, J. B., (1995) The monetary transmission mechanism: An empirical framework.
Journal of Economic Perspectives, 9 (4), pp. 11-26.
Townsend, R., (1979) Optimal contracts and competitive markets with costly state veri-
fication. Journal of Economic Theory, 21 (2). pp - 265-293.
Turner, P., (2012) Weathering financial crisis: domestic bond markets in EMEs. BIS
Papers chapters,In: Bank for International Settlements (ed.), Weathering financial crises:
bond markets in Asia and the Pacific, 63, pp. 15-34.
RBI (2014): Report of the expert committee to revise and strengthen the monetary policy
framework, Reserve Bank of India.
Van den Heuvel, S. J., (2002) Does bank capital matter for monetary transmission?
Economic Policy Review, Federal Reserve Bank of New York, pp. 259-265, May.
Vlcek, J. , & Roger, S., (2012) Macrofinancial Modeling At Central Banks: Recent
Developments and Future Directions. IMF Working Papers 12/21, International Monetary
Fund.
Woodford, M., (2003) Interest and prices foundations of a theory of monetary policy.
Princeton: Princeton University Press.
67
7 Appendix A: Alternative Channels of Monetary Trans-
mission
As mentioned in Section 2.1, there are six different channels of monetary transmission pro-
posed in the literature. We provide a brief summary of these channels in the following
text.
The first one is the conventional money channel, through which monetary tightening
raises the short-term interest rates, changes the composition of demand for monetary and
non-monetary financial assets, reduces the demand for investment and impact on output
growth and inflation in the economy (Kletzer, 2012).
Second, the Keynesian view of interest rate channel of transmission mechanism suggests
that a policy-induced increase in the short-term nominal interest rate leads to an increase
in the longer-term nominal interest rates. The reason is investors act to arbitrage away the
differences in the risk-adjusted expected returns on debt instruments of different maturities
as described by the expectations hypothesis of the term structure. Given the nominal rigidity
in place, the movements in nominal interest rates sluggishly translate into movements in real
interest rates. Such changes in real interest rates alter the relative price of future consumption
to current consumption, user cost of capital for fixed investment, relative price of domestic
goods in terms of foreign goods, and impact on economic activity in the economy.
Third, according to the credit channel of monetary transmission, design of financial
contract departs from the Modigliani-Miller axioms under the circumstances of market in-
completeness and heterogeneity among the economic agents, and leads to various kinds of
frictions. These frictions give rise of a wedge between the cost of capital incurred by the
capital users and the return from capital received by the capital providers. Due to presence
of such wedge, changes in the monetary policy instruments can create its impact on the
real activity by changing the price of credit or credit limits in the credit market with an
additional strength. It does so through two ways: (i) the balance sheet channel, in which
monetary policy affects borrowers’net worth and debt collateral, and (ii) the bank lending
channel, in which policy instrument impacts on the level of intermediated credit. These
channels have been widely incorporated into general equilibrium models and found to be
empirically relevant to explain the propagation mechanism of the policy shocks.24
Fourth, in the open economies under flexible exchange rate framework, exchange rate
channel operates through capital flows in response to policy interest rate changes. A key
24Evidence on credit channel of monetary policy transmission can be found in Fazzari et al. (1988),Bernanke and Blinder (1992), Kashyap et al. (1993), and Bernanke, Gertler, and Gilchrist (1999) foradvanced countries, and in Mohanty and Turner (2008), Agenor and Pereira da Silva , (2014) for the EMDEs.
68
assumption underpinning this relationship is the Uncovered Interest rate Parity (UIP) condi-
tion, which suggests that a policy-led cut in the interest rate makes domestic interest bearing
assets less attractive vis-à-vis foreign assets and leads to capital outflows. This entails to
depreciation of the nominal exchange rate. In a sticky-price environment, it turns into a
real depreciation and an increase in the price of tradeable relative to non-tradeable. Finally,
it impacts on output and inflation, though with a lag, depending on the share of imported
goods in consumption basket. This exchange rate channel can trigger first and second round
effects on inflation via the cost of production as many intermediate goods coming through
imports may become expensive.25
Fifth, the asset price channel works through the impact of monetary policy on the do-
mestic asset prices which include bond, stock and real estate prices. Changes in the policy
interest rate affect the households and the firms through valuation of equities. Higher interest
rates bring down the equity prices, reduces the present value of financial wealth, and leaves
contractionary impact on consumption of the economic agents. Besides, the drop in equity
prices also impacts on the Tobin’s Q. Low value of the equities relative to the replacement
cost of capital results in reduced investment spending in the economy.26
Sixth, changes in the policy rate can influence the time path of real activity through
the expectation channel and the confidence with which those expectations are held. Such
changes in perception will affect participants in the commodity, labour and financial markets.
According to Blinder (2000), a successful monetary policy emerges as a result of the effective
management of expectations instead of the policy rate alone. Depending on the perception
of households and firms about intertemporal rates of substitution, the expectation channel
can transmit the impact of monetary policy shocks. Mohanty and Turner (2008) argued that
depending on the degree of credibility, predictability of actions, and commitment to vary the
policy instrument consistently by the central bank, the expectation channel contributes to
speed up the adjustment and impacts the transmission lags significantly.
25Since 2010, the problem of currency mismatches in many emerging markets has increased — notablybecause of a substantial increase in the foreign currency borrowing by the non-financial companies (Chuiet al, 2016). This has made the exchange rate channel more relevant in the monetary policy debate of thepost global financial crisis. The channel has magnified for both large commodity exporting countries suchas Brazil and Russia, as well as the commodity importing countries like India, Indonesia and Thailand.26The empirical evidence of many advanced and emerging market economies suggests that monetary policy
and asset prices had some relationship, especially pertaining to house prices and stock market movements.However, this channel has weakened substantially in the post-financial crisis period (Turner, 2012).
69
8 Appendix B: Econometric Specification of SVARAnaly-
sis
The system of equations representing SVAR framework can be written in vector form as:
B0xt = A0 +B1xt−1 +B2xt−2 + ...+Bpxt−p + ut (75)
where, xt is an n × 1 vector of endogenous variables, A is an n × 1 vector of constant
terms, and ut is an n×1 vector of structural disturbances, and p denotes the number of lags.
In the reduced form, it can be rewritten as:
xt = A1 + Φ1xt−1 + Φ2xt−2 + ...+ Φpxt−p + et (76)
where, et is the reduced form residuals and equals to B−10 ut.
For our analysis, we have chosen: xt =[yt πt bt i
dt i
bt it]′where, yt is real GDP at factor
cost, πt is CPI inflation, bt denotes non-food credit growth, idt stands for deposit rate, ibt
lending rate and call money rate is denoted by it.
The relationship between ut and et :
ut = B0et (77)
Based on the above relation, the scheme of identification is summarised below:
uyt
uπt
ubt
uid
t
uib
t
uit
=
1 0 0 0 0 0
b21 1 0 0 0 0
b31 b32 1 0 b35 0
b41 b42 b43 1 0 0
b51 b52 b53 b54 1 0
0 0 0 0 0 1
eyt
eπt
ebt
eid
t
eib
t
eit
(78)
where, ut is the vector of structural innovations, et is the vector of reduced form residuals
and bij represents contemporaneous structural parameters.
70
9 Appendix C: Data Sources and Computations
Quarterly data on real sector variables measured at constant prices with base 2011-12, viz.
Gross Value Added (GVA), Gross Domestic Products (GDP), Private Final Consumption
Expenditure, Government Final Consumption Expenditure and Gross Fixed Capital Forma-
tion have been extracted from the Central Statistics Offi ce (CSO), Government of India. As
the quarterly data for the new series with base 2011-12 are available from 2011-12: Q1, the
back series prior to 2011-12 are derived by splicing.
The quarterly data for the price index is derived as the average of the respective monthly
price index during the quarter. Data on Consumer Price Index (CPI) with base year at 2012
are collected from the website of CSO, Government of India. As the historical data on CPI
are not available prior to the year 2011, the same are spliced using the CPI for Industrial
Workers (CPI-IW), published by the Indian Labour Bureau, Government of India.
Data on banking and interest rate variables such as bank credit, non-food credit, repo
rate, call money rate, cash reserve ratio, SLR, deposit rates and weighted average lending
rates were collected from the website of the RBI. The RBI has been disseminating the
monthly Weighted Average Lending Rate (WALR) on outstanding loan from Feb-12. For
compilation of lending rates prior to Feb-12, the end quarter interest rate (Term loan interest
rate other than export credit) for five scheduled commercial banks has been used.27 The
simple average of the maximum and minimum lending rate for each of the banks has been
considered and the lending rate is obtained as the weighted average of each these banks
interest rate, with weights being proportion to the average outstanding credit amount as on
end 2010-11 and 2011-12. As a proxy of investment in housing sector, housing price data
are taken from the RBI database. The steady-state share of bank deposit to GDP ratio is
obtained from the database of St. Louise FRED (https://fred.stlouisfed.org).
The quarterly Net Interest Margin (NIM) was derived using the Chow-Lin method from
annual NIM data. The annual NIM data were collected from the annual publication Statis-
tical Tables elating to Banks in India, published by the RBI. The current and lag quarters
of nominal GDP was used to derive the quarterly NIM estimates.
27The banks selected for this purpose includes three nationalised banks (State Bank of India, Bank ofBaroda and Punjab National Bank), one private bank (ICICI Bank Ltd.) and one foreign bank (Citibank).
71
10 Appendix D: Log-linearised Model and Description
of Steady-state
10.1 Liquidity-constrained Household
CR,t + σLLR,t = ωR,t (79)
CR,t = ωR,t + LR,t (80)
10.2 Patient Household
λP,t = −(
1
1− σh
)(CP,t − σhCP,t−1
)(81)
LP,t =
(1
σL
)(ωP,t + λP,t
)(82)
εH,t − HP,t =
{1
1− βP (1− δh)
}(λP,t + Qh
t
)−{
βP (1− δh)1− βP (1− δh)
}(λP,t+1 + Qh
t+1
)(83)
λP,t =(ıdt − πt+1
)+ λP,t+1 (84)
{(CPC
)(C
Y
)}CP,t +
{Qh
(HP
H
)(H
Y
)}[(Qht + HP,t
)− (1− δh)
(Qht + HP,t−1
)]+
(D
Y
)Dt +
(TXP
TX
)(TX
Y
)TXP,t
=
{(1− γR) γ (1− α)
X
}(ωP,t + LP,t
)+
{(1 + id
π
)(D
Y
)}(ıdt−1 − πt + Dt−1
)(85)
10.3 Impatient Household
λI,t = −(
1
1− σh
)(CI,t − σhCI,t−1
)(86)
LI,t =
(1
σL
)(ωI,t + λI,t
)(87)
72
εH,t − HI,t = $H
(λI,t + Qh
t
)− {βI (1− δh)$H}
(λI,t+1 + Qh
t+1
)−[εHLV (1− δh)
{π − βI
(1 + ibH
)}$H
π (1 + ibH)
](µI,t + Qh
t+1 + εHLV,t
)(88)
where, $H =(λIHIεH
)=
[{1− βI (1− δh)} −
{εHLV (1−δh){π−βI(1+ibH)}
π(1+ibH)
}]−1
λI,t =
{βI(1 + ibH
)π
}(ıbHt − πt+1 + λI,t+1
)+
{1−
βI(1 + ibH
)π
}(ıbHt + µI,t
)(89)
{(CIC
)(C
Y
)}CI,t +
{Qh
(HI
H
)(H
Y
)}[(Qht + HI,t
)− (1− δh)
(Qht + HI,t−1
)]+
{(1 + ibH
π
)(BH
B
)(B
Y
)}(ıbHt−1 − πt + BH,t−1
)=
{(1− γR) (1− γ) (1− α)
X
}(ωI,t + LI,t
)+
{(BH
B
)(B
Y
)}BH,t (90)
(ıbHt + BH,t
)= εHLV,t + πt+1 + Qh
t+1 + HI,t (91)
10.4 Wholesale Goods Producing Entrepreneur
λE,t = −(
1
1− σh
)(CE,t − σhCE,t−1
)(92)
YE,t = εA,t + α(ut + Kt−1
)+ (1− α) Lt (93)
rkt = YE,t − xt − ut − Kt−1 (94)
rkt =
(ψbψa
)ut (95)
ωR,t = YE,t − Xt − LR,t (96)
73
ωP,t = YE,t − Xt − LP,t (97)
ωI,t = YE,t − Xt − LI,t (98)
λE,t =
{βE
(1 + ibE
π
)}(ıbEt − πt+1 + λE,t+1
)+
{1− βE
(1 + ibE
π
)}(ıbEt + µE,t
)(99)
(λE,t + Qk
t
)= [1− βE {ψa + (1− δk)}]
(µE,t + εELV,t + Qk
t+1 + πt+1
)+ (βEψa)
(λE,t+1 + rkt+1 + ut+1
)+ {βE (1− δk)}
(λE,t+1 + Qk
t+1
)− βE
(λE,t+1 + ψaut+1
)(100)
{(CEC
)(C
Y
)}CE,t +
(Ik
δkY
)(Qkt + Kt
)−{(
1− δkδk
)(Ik
Y
)}(Qkt + Kt−1
)+
{(1 + ibE
π
)(BE
B
)(B
Y
)}(ıbEt−1 − πt + BE,t−1
)+
(ψaI
k
δkY
)ut
=( αX
)(YE,t − Xt
)+
{(BE
B
)(B
Y
)}BE,t (101)
(ıbEt + BE,t
)= εELV,t + πt+1 + Qk
t+1 + Kt (102)
10.5 Final Goods Producing Retailer
πt =
{−εy
Xϑpπ2 (1 + βP θp)
}Xt +
(βP
1 + βP θp
)πt+1 +
(θp
1 + βP θp
)πt−1 (103)
10.6 Capital Goods and Housing Goods Producing Sectors
Ikt =
(%k
1 + βP
)Qkt +
(βP
1 + βP
)Ikt+1 +
(1
1 + βP
)Ikt−1 −
(βP
1 + βP
)(εik,t+1 − εik,t
)(104)
Iht =
(%h
1 + βP
)Qht +
(βP
1 + βP
)Iht+1 +
(1
1 + βP
)Iht−1 (105)
74
Kt = (1− δk) Kt−1 + δkIkt (106)
Ht = (1− δh) Ht−1 + δhIht (107)
10.7 Retail Banking Sector Operations
ıdt = ıst −{
φd(εd + 1) id
}(ıdt − ıdt−1
)+
{βPφd
(εd + 1) id
}(ıdt+1 − ıdt
)(108)
ıbHt = ıbt −{
φbH(εbH − 1) ibH
}(ıbHt − ıbHt−1
)+
{βPφbH
(εbH − 1) ibH
}(ıbHt+1 − ıbHt
)(109)
ıbEt = ıbt −{
φbE(εbE − 1) ibE
}(ıbEt − ıbEt−1
)+
{βPφbE
(εbE − 1) ibE
}(ıbEt+1 − ıbEt
)(110)
10.8 Wholesale Banking Sector Operations
ıst =
{(1− αc − αs)
(i
is
)}ıt +
(αsi
g
is
)ıgt (111)
ıbt = ıt − φzκ3b(Zt − Bt
)(112)
Bt = (1− αc − αs) Dt +
(BIB
B
)BIBt +
(Z
B
)Zt (113)
πt + Zt =
(1− δbπ
)Zt +
{1−
(1− δbπ
)}Πbt−1 (114)
Πbt =
{ibH
κb (π − 1 + δb)
(BH
B
)}(ıbHt + BH,t
)+
{ibE
κb (π − 1 + δb)
(BE
B
)}(ıbEt + BE,t
)+
{αsi
g
κb (π − 1 + δb)
}(ıgt + Dt
)−{
id
κb (π − 1 + δb)
}(ıdt + Dt
)−{
i
κb (π − 1 + δb)
(BIB
B
)}(ıt + BIB
t
)(115)
10.9 Fiscal Policy Block(G
Y
)Gt +
{(1 + ig
π
)αs
(D
Y
)}(ıgt−1 − πt + Dt−1
)=
(TX
Y
)TX t +
{αs
(D
Y
)}Dt
(116)
75
10.10 Monetary Policy Block
ıt = ϕiıt−1 + (1− ϕi)[ϕyYt + ϕπEt {πt+1}
]+ εm,t (117)
10.11 Aggregation & Market Clearing Conditions
Ct =
(CRC
)CR,t +
(CPC
)CP,t +
(CIC
)CI,t +
(CEC
)CE,t (118)
Lt = γRLR,t + (1− γR) γLP,t + (1− γR) (1− γ) LI,t (119)
ωt = γRωR,t + (1− γR) γωP,t + (1− γR) (1− γ) ωI,t (120)
Ht =
(HP
H
)HP,t +
(HI
H
)HI,t (121)
Bt =
(BH
B
)BH,t +
(BE
B
)BE,t (122)
Bt = Dt (123)
Yt =
(C
Y
)Ct+
(Ik
Y
)Ikt +
(Ih
Y
)Iht +
(G
Y
)Gt+{(
ψaδk
)(Ik
Y
)}ut+
{(δbκbπ
)(B
Y
)}(Zt−1 − πt
)(124)
10.12 List of Shock Variables
The set of shock variables{εA,t, εik,t, εm,t, Gt, εh,t, ε
HLV,t, ε
ELV,t, εy,t
}∞t=0
, incorporated in the
model, follow AR(1) process.
10.13 Description of Steady-state(BH
B
)=
{1−
(BE
B
)}(125)
(HI
H
)=
(1
εELV
)(δh
1− δh
){(B/Y )
(Ik/Y
}(1 + ibH
π
)(BH
B
)(126)
76
(HP
H
)= 1−
(HI
H
)(127)
(CEC
)=
1
(C/Y )
[( αX
)−(
1 + ibE − ππ
)(BE
B
)(B
Y
)−(Ik
Y
)](128)
(CIC
)=
1
(C/Y )
{(1−γR)(1−γ)(1−α)
X
}−(
1+ibH−ππ
) (BHB
) (BY
)−(HIH
) (Ih
Y
) (129)
(CPC
)=
1
(C/Y )
[{(1− γR) γ (1− α)
X
}+
(1 + id − π
π
)(D
Y
)−(HP
H
)(Ih
Y
)](130)
(CRC
)=
{1−
(CPC
)−(CIC
)−(CEC
)}(131)
77
11 Appendix E: Figures on Counterfactual Experiments
.006
.004
.002
.000
.002
.004
.006
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ID_LTV_H_HIGH ID_LTV_H_LOW
.004
.000
.004
.008
.012
.016
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IL_LTV_H_HIGH IL_LTV_H_LOW
.20
.16
.12
.08
.04
.00
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B_LTV_H_HIGH B_LTV_H_LOW
.006
.004
.002
.000
.002
.004
.006
.008
.010
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
C_LTV_H_HIGH C_LTV_H_LOW
.028
.024
.020
.016
.012
.008
.004
.000
.004
.008
.012
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IK_LTV_H_HIGH IK_LTV_H_LOW
.10
.08
.06
.04
.02
.00
.02
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IH_LTV_H_HIGH IH_LTV_H_LOW
.030
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
L_LTV_H_HIGH L_LTV_H_LOW
.004
.003
.002
.001
.000
.001
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
PI_LTV_H_HIGH PI_LTV_H_LOW
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Y_LTV_H_HIGH Y_LTV_H_LOW
Figure 13: Effects of Change in Collateral Constraint for Households (LTV)
.006
.004
.002
.000
.002
.004
.006
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ID_GAMMA_R_HIGH ID_GAMMA_R_LOW
.004
.000
.004
.008
.012
.016
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IL_GAMMA_R_HIGH IL_GAMMA_R_LOW
.12
.10
.08
.06
.04
.02
.00
.02
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B_GAMMA_R_HIGH B_GAMMA_R_LOW
.006
.004
.002
.000
.002
.004
.006
.008
.010
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
C_GAMMA_R_HIGH C_GAMMA_R_LOW
.028
.024
.020
.016
.012
.008
.004
.000
.004
.008
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IK_GAMMA_R_HIGH IK_GAMMA_R_LOW
.10
.08
.06
.04
.02
.00
.02
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IH_GAMMA_R_HIGH IH_GAMMA_R_LOW
.024
.020
.016
.012
.008
.004
.000
.004
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
L_GAMMA_R_HIGH L_GAMMA_R_LOW
.004
.003
.002
.001
.000
.001
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
PI_GAMMA_R_HIGH PI_GAMMA_R_LOW
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Y_GAMMA_R_HIGH Y_GAMMA_R_LOW
Figure 14: Effects of Change in Proportion of Liquidity Constrained Household
78
.006
.004
.002
.000
.002
.004
.006
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ID_GAMMA_HIGH ID_GAMMA_LOW
.0050
.0025
.0000
.0025
.0050
.0075
.0100
.0125
.0150
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IL_GAMMA_HIGH IL_GAMMA_LOW
.12
.10
.08
.06
.04
.02
.00
.02
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B_GAMMA_HIGH B_GAMMA_LOW
.006
.004
.002
.000
.002
.004
.006
.008
.010
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
C_GAMMA_HIGH C_GAMMA_LOW
.030
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IK_GAMMA_HIGH IK_GAMMA_LOW
.08
.06
.04
.02
.00
.02
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IH_GAMMA_HIGH IH_GAMMA_LOW
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
L_GAMMA_HIGH L_GAMMA_LOW
.005
.004
.003
.002
.001
.000
.001
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
PI_GAMMA_HIGH PI_GAMMA_LOW
.020
.016
.012
.008
.004
.000
.004
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Y_GAMMA_HIGH Y_GAMMA_LOW
Figure 15: Effects of Change in Proportion of Patient Household
.006
.004
.002
.000
.002
.004
.006
.008
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ID_PHI_Z_HIGH ID_PHI_Z_LOW
.0050
.0025
.0000
.0025
.0050
.0075
.0100
.0125
.0150
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IL_PHI_Z_HIGH IL_PHI_Z_LOW
.12
.10
.08
.06
.04
.02
.00
.02
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B_PHI_Z_HIGH B_PHI_Z_LOW
.006
.004
.002
.000
.002
.004
.006
.008
.010
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
C_PHI_Z_HIGH C_PHI_Z_LOW
.028
.024
.020
.016
.012
.008
.004
.000
.004
.008
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IK_PHI_Z_HIGH IK_PHI_Z_LOW
.08
.06
.04
.02
.00
.02
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IH_PHI_Z_HIGH IH_PHI_Z_LOW
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
L_PHI_Z_HIGH L_PHI_Z_LOW
.005
.004
.003
.002
.001
.000
.001
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
PI_PHI_Z_HIGH PI_PHI_Z_LOW
.024
.020
.016
.012
.008
.004
.000
.004
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Y_PHI_Z_HIGH Y_PHI_Z_LOW
Figure 16: Effects of Change in Bank Capital Adjustment Cost
79
.008
.006
.004
.002
.000
.002
.004
.006
.008
.010
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
ID_KAPA_B_HIGH ID_KAPA_B_LOW
.010
.005
.000
.005
.010
.015
.020
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IL_KAPA_B_HIGH IL_KAPA_B_LOW
.20
.16
.12
.08
.04
.00
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B_KAPA_B_HIGH B_KAPA_B_LOW
.008
.004
.000
.004
.008
.012
.016
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
C_KAPA_B_HIGH C_KAPA_B_LOW
.05
.04
.03
.02
.01
.00
.01
.02
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IK_KAPA_B_HIGH IK_KAPA_B_LOW
.12
.10
.08
.06
.04
.02
.00
.02
.04
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
IH_KAPA_B_HIGH IH_KAPA_B_LOW
.030
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
L_KAPA_B_HIGH L_KAPA_B_LOW
.006
.005
.004
.003
.002
.001
.000
.001
.002
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
PI_KAPA_B_HIGH PI_KAPA_B_LOW
.030
.025
.020
.015
.010
.005
.000
.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Y_KAPA_B_HIGH Y_KAPA_B_LOW
Figure 17: Effects of Change in Capital Adequacy Ratio
80